U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS OFFICE OF PUNNING September 1965 Digitized by the Internet Archive in 2012 with funding from LYRASIS Members and Sloan Foundation http://archive.org/details/trafficassignmenOOunit TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS U.S. DEPARTMENT OF COMMERCE John T. Connor, Secretary BUREAU OF PUBLIC ROADS Rex M. Whitton, Administrator For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C., 20402 - Price $1 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS PREFACE September 1965 The agencies concerned with making decisions on when and where to construct new urban transportation facilities, or to improve existing ones, must consider many factors in reaching sound decisions. Consequently, considerable time, money, and effort are usually expended in the planning and design of such improvements. For the most part, the planning work is carried on within the framework of an urban transportation study. Since the end of World War II, a number of urban transportation studies have been conducted in an increasingly comprehensive manner. Significant improve- ments in both basic study philosophy and analytical methodology have greatly contributed to a better understanding of the urban transportation problem. This understanding has resulted in a concerted effort toward the development of a transportation planning process which utilizes the interrelationships between available transportation facilities, land use, and socio-economic characteristics of the population to provide quantitative information on the travel demands created by alternate land use patterns and transportation systems in an urban area. Such information can then be used by various agencies to make decisions about when and where improvements should be made in transportation networks to satisfy the present and future travel demands and to promote desirable land development patterns. The technical considerations of an urban transportation study generally involve four integrated phases; (l) data collection; (2) data analysis and estimating; (3) policy decision making; and (k) plan implementation. The integration of the four phases is achieved by designing the data-gathering phase to obtain the facts necessary for developing estimating procedures for making estimates. These estimating procedures must provide quantitative information about the probable consequences of decisions concerning alternate urban development and transportation plans. Decisions based upon a knowledge of the alternate plans will in turn permit the development of programs for plan implementation. Briefly, the four phases accomplish the following: • The data collection phase consists of the origin-destination survey and its associated inventories. • The data analysis and estimating phase, using the data collected in the first phase, undertakes the population analysis, economic analysis, land use analysis, trip generation estimates, trip distribution estimates, and traffic assignments. • The traffic assignments are used to provide information on the capabilities of the various alternate transportation systems. • The plan implementation phase develops from the decision making phase as a set of construction priorities based on criteria which include such factors as community interest, availability of funds, limitations on revenue, present programs and the public works plans of other jurisdictions. This manual covers the details and mechanics of computer application for traffic assignment and trip distribution. The remainder of this preface describes the contents of this manual in greater and more technical detail. Traffic assignment and trip distribution are two key phases of the trans- portation planning process. These phases generate the quantitative data on travel needed to properly plan transportation facilities. The traffic assignment techniques provide an estimate of the probable traffic on each segment of a transportation network. The alternate methods of assignment to be covered in this volume are the all-or-nothing and diversion methods. The history and theory of traffic assignment are covered in detail in a previous publication by the Bureau of Public Roads(l)i/. Trip distribution techniques provide for distributing the trips emanating from each zone in the study area to other zones. This manual covers the necessary phases for trip distribution with the gravity model formula or the Fratar formula. The history and theory of the gravity model are covered in reference (2). The system described in this volume is designed to process a basic set of trip data cards. It allows the computation of surveyed trip length frequency, distribution of trips between zones by the gravity model formula, and assignment of trip interchanges to an existing transportation network. The gravity model program automatically adjusts zonal attractions for a given set of traveltime factors and computes the new trip length frequency from which traveltime factors can be adjusted. The system also accommodates estimates of future trip production and trip attraction values for use in a gravity model trip distribution, and then goes on to assign the trip distribution to a future transportation network. The analytical procedures described in this volume are mechanized by a highly flexible series of computer programs which utilize the IBM 1620 (60K) electronic computer, coupled with the capabilities of IBM electronic accounting machines. This battery of computer programs was developed by several State highway departments and the Bureau of Public Roads, and is generally applicable to urban areas with populations up to 150,000 persons. In order to utilize the computer programs for this system, the study area must be described by a maximum of 699 nodes, of which 200 may be zone centroids. The maximum traveltime that can be accommodated is 99 minutes. 1/ The numbers in parentheses identify references in appendix A. U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS TABLE OF CONTENTS September 1965 Page PREFACE i CHAPTER I. INTRODUCTION A. The Process 1-1 B. Traffic Assignment 1-1 C. Trip Distribution , 1-3 CHAPTER II. INVENTORIES AND DECISIONS A. General II-l B. Travel Pattern Inventory II-l C. Travel Facilities Inventory II-l D. Socio -Economic and Land Use Data II -2 E. Initial Decisions to be Made II -3 1. Vehicle trip ADT model II -3 2. Peak hour' or off-peak hour driving time II -3 3. Trip purpose classification II -3 a. Home based work trips II -k b. Home based nonwork trips II -k c. Nonhome based trips II -k k. Treatment of external trips 11-^4- CHAPTER III . DESCRIBING THE EXISTING SYSTEM AND ASSIGNING TRAFFIC A. Preparation of the Network III-l 1. Map preparation III-l 2. Speed and traveltime information III-l 3. Traffic volumes Ill -2 h. Street capacities Ill -2 5. The network description Ill -2 September 1965 Page Locate and number the zones and centroids . . . Ill -2 Define the basic transportation network .... Ill -3 Connect the centroids Ill -3 Locate and define the nodes Ill -7 Prepare a list of nodes Ill -8 Assign the node numbers Ill -10 Set turn penalties and turn prohibitors .... Ill -10 Define the link parameters Ill -11 6. Special treatment for external stations Ill -11 7- Identifying and storing the maps and tabulations. . Ill -12 8. Link data card preparation Ill -12 9. Checking for errors Ill -12 B. Editing the Network *. . III-13 C. Building and Checking Minimum Time Path Trees Ill -15 D. Processing the survey trip cards Ill -20 Selecting trip cards Ill -20 Editing trip records Ill -20 Sorting and linking trip cards Ill -23 Separating trip cards by purpose Ill -26 Building the trip tables Ill -27 E. Assigning 0-D Trips to the Present Network. CHAPTER IV. THE GRAVITY MODEL III -27 The Gravity Model Formula IV-1 Definitions of Parameters IV -2 Trip production and attraction IV-2 Spatial separation between zones TV-3 Traveltime factors TV-3 Zone -to -zone adjustment factors IV -k C. Data Needed to Determine Parameters IV -k CHAPTER V. CALIBRATING THE GRAVITY MODEL A. General V-l B. Phase One --Processing Basic Data V-2 1. Selecting trip records V-2 2. Determining interzonal driving time V-2 3. Determining intrazonal driving time V-2 September 1965 k. Determining the traveltimes V-2 5. Updating the skim trees V-3 C. Phase Two --Obtaining Productions and Attractions V-4 D. Phase Three - -Calibrating the Gravity Model V-4 1. Developing 0-D trip length frequency distributions. . . V-4 2. Selecting initial traveltime factors V-6 3- Iteration procedure to get final traveltime factors . . V-6 E. Phase Four --Balancing Attractions for the Final Calibration V-12 F. Phase Five --Adjustment Factors for the Gravity Model. . . V-12 1. Topographical barriers V-12 2. Developing zone-to-zone adjustment factors V-lk CHAPTER VI- ASSIGNING AND TESTING THE GRAVITY MODEL TRIP DISTRIBUTION A. General VI -1 B. Assigning the Gravity Model Trips to the Present Network. VT-1 C. Statistical Tests VI -1 D. Analytical Tests VI -3 E . Converting the Gravity Model Results to Directional Movements VI -6 CHAPTER VII. PREPARING THE PROPOSED STREET AND HIGHWAY NETWORK A. Coding the Network VII -1 B. Editing the Proposed System Network VII -2 C. Building and Checking Proposed System Trees VII -2 D. Alternate Proposed Systems VII -k CHAPTER VIII. DISTRIBUTION OF FUTURE TRIPS WITH THE GRAVITY MODEL A. General VIII-1 B. Future Trip Generation VIII -2 C. The Gravity Model Distribution for a Proposed System. . VIII -5 1. Productions and attractions VIII -5 2. Skim trees VIII -5 3- Traveltimes and traveltime factors VIII -5 CHAPTER IX. THE FRATAR TRIP DISTRIBUTION A. The Fratar Theory IX -1 Contents September 1965 Page B. Using the Fratar Procedure IX -2 C. Data Requirements IX -2 1. Trip tables IX -2 2. Growth factors IX -2 3. Precautionary measures IX -3 D. Program Options. IX -3 E. Checking the Trip Distribution IX -3 CHAPTER X. ANALYSIS OF PROPOSED SYSTEMS A. Assignment of Future Traffic to Proposed Systems X-l B. Analysis of Systems X-l 1. Testing X-2 2. Analysis X-2 3. Evaluation X-3 k. Adjustments X-3 C. Accuracy of Traffic Assignments X-k CHAPTER XI. ASSIGNMENT 1 PROGRAM XI -1 CHAPTER XII. TRIP TABIE BUILDER PROGRAM XII -1 CHAPTER XIII. ADD TRIP TABIES PROGRAM XIII -1 CHAPTER XIV. ASSIGNMENT 2 PROGRAM XIV-1 CHAPTER XV. UPDATE SKIM TREES PROGRAM XV-1 CHAPTER XVI. TRIP IENGTH FREQUENCY PROGRAM XVI -1 CHAPTER XVII. TRIP END SUMMARY PROGRAM XVII -1 CHAPTER XVIII. GRAVITY MODEL PROGRAM XVIII -1 CHAPTER XIX. TRIP COMPARISON PROGRAM XIX-1 CHAPTER XX. FRATAR TRIP DISTRIBUTION PROGRAM XX-1 APPENDIX A REFERENCES A-l U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER I - INTRODUCTION September 1965 A. The Process It is now widely recognized that urban travel patterns are a function of: 1. The type and extent of transportation facilities in an area; 2. The pattern of activities in an area, including the location and intensity of land use; and 3. The social and economic characteristics of the population of an urban area. The recognition of these facts has resulted in the development of a travel forecasting process which utilizes the above relationships to provide infor- mation on the travel demands created by the distribution of activities and the transportation system. The technical considerations of the urban travel forecasting process involve four integrated steps: (l) the inventories; (2) the analysis of existing conditions and calibration of forecasting techniques; (3) the forecast; and (k) systems analysis. The complete process is shown in figure 1-1; the cross-hatched areas of the diagram will be discussed in this manual. In order to present their proper positions in the urban travel forecasting process, different phases of this process will be discussed in turn, with the traffic assignment and trip distribution phases being treated in more detail. A brief discussion of trip generation is provided in chapter VIII. Before getting into the details of each phase, some general statements are in order. A statement about traffic assignment follows in section B. and a statement about trip distribution is in section C. B. Traffic Assignment The traffic assignment phase refers to the estimation of the number of vehicles that will use each individual section of the transportation system. The estimation of the load on various sections or links may be for present conditions or any future year. For future year assignments, a travel forecast must first be made in order to obtain the volumes. After forecasts of the future travel have been made, the volumes are assigned to a transportation network. The results are evaluated in accordance with the desired level of service and the social and economic consequences of the system. The 1-1 URBAN TRAVEL FORECASTING PROCESS INVENTORIES ECONOMIC A AND POPULATION >- ANALYSIS OF EXISTING CONDITIONS AND CALIBRATION OF ^ FORECASTING TECHNIQUES ECONOMIC ACTIVITY, AND POPULATION PROJECTION TECHNIQUES FORECAST < TECHNIQUES POPULATION TRIP ACCURAC .ERATION +-///* mPXMu^/. ^LEtTI^NO^ TWORKAND X T INITIAL ASsfc-NMENt // /tw / / / ' NETWORK AI , M °^ L / FEEDBACK ( IMPLEMENTATION FUTIH •RIP DISTRIBUTION I / ASSIGNMENT / z Figure I -1. --Urban Travel Forecasting Process. resulting revisions are made to the system and additonal future assignments are made to the revised networks. All or part of the traffic assignment and distribution phases are repeated until satisfactory results have been obtained. C. Trip Distribution A key element of the analysis and forecasting phase is the development of a reliable procedure that is capable of estimating zonal trip interchanges for alternative configurations of land use and transportation facilities. These zonal interchanges constitute a basic part of the travel information necessary for transportation planning. The analysis of many origin and destination surveys in relation to the type and extent of the transportation facilities available, the uses of land, and the various social and economic characteristics of trip makers, indicates that zonal trip interchanges can be estimated by the application of mathematical formulas, called "trip distribution models." The use of such models in transportation planning offers certain advantages over older trip distribution forecasting techniques. Earlier it was suggested that a basic aim of the transportation planning process is to provide decision makers with quantitative information about the consequences of decisions concerning the type, location, size, and timing of transportation improvements. One of the advantages of mathematical travel models is that they serve this aim quite well. Mathematical trip distribution models provide a common base for simulating the travel patterns that can be expected to result from a variety of different land use configurations coupled with alternate highway systems. This allows the responsible authorities to see, in advance, the probable consequences, in terms of travel patterns, of different combinations of land use and transportation patterns. 1-3 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER II - INVENTORIES AND DECISIONS September I965 A. General The following sections discuss the information vhich must be available and the decisions vhich must be made concerning the data from the various inven- tories so that it will provide the required results when used in the various phases of the planning process. B. Travel Pattern Inventory The procedures described here assume that origin and destination survey data are available in sufficient amounts to provide the following information: 1. Zone-to-zone movements by trip purpose 2. Productions and attractions In these respects, the standard home interview survey, conducted in a manner as recommended by the Bureau of Public Roads (3), yields the most complete and accurate data. In all cases, sufficient sampling controls must be exercised to assure that the resulting travel data are complete and not statistically biased in any manner. In addition to a home interview survey of internal travel, an external cordon survey (6) and a truck and taxi survey (7) should also be included to complete the picture of travel in the urban area under study. Other special surveys may be necessary, depending on the area under consideration. C. Travel Facilities Inventory Besides sufficient, reliable data on travel patterns, there must also be available several items of information relating to the travel facilities in the area. For the purposes of traffic assignment and gravity model trip distribution, the following data on major sections of the highway and transit facilities in the area are particularly important: 1 . Location 2. Physical dimensions a. Length b. Width c. Number of lanes 3. Average speed of travel in both peak and off-peak hours k. Signal! zation, parking requirements, direction of travel and other data for capacity calculations 5. Existing traffic volumes D. Socio -Economic and Land Use Data In order to provide an effective analysis of travel patterns and trip generation the following data should be available on a zonal basis: 1. Population by age group 2. Average family income 3. Labor force i/ k. Employment by industry and occupation £/ 5. Land area by type (8) 6. Car ownership 7. Number of dwelling units These data will provide the information for estimating trip production and trip attraction values. In the remaining discussions it will be assumed that all the required data have been obtained and are available for use. l/ "Labor force" refers to workers considered in relationship to their place of residence. 2/ "Employment" refers to workers considered in relationship to their place of work. E. Initial Decisions to be Made 1. Vehicle trip APT model. — Once a transportation study has decided on the trip distribution model to be used, in this case the gravity model, there remain many choices as to the manner in which this model is to be used to provide estimates of travel patterns. Perhaps the first choice is whether the model should distribute total person trips or vehicle trips. The answer to this question is directly related to both the objectives and needs of the study and to the size of the area involved. In small urban areas, where transit usage is not or will not be very large, the model should be designed for vehicular trips. The procedures described herein can be used to distribute either person or vehicular trips, as desired by the user. In large urban areas it is generally necessary to analyze different levels of both highway and public transit service in order to arrive at a properly balanced transportation system. For this reason a person trip model is sometimes used. When person trips are distributed between zones either in total or by separate modes of travel, depending on the procedure desired, it is necessary to take into account the modal split and vehicle occupancy rates of these trips prior to assigning them to the transportation system. 2. Peak hour or off-peak hour driving time . --Still another decision must be made concerning the time to be used to indicate the driving time portion of the spatial separation between zones. The selection of peak, off-peak or some combination of these two times depends primarily on the existing congestion conditions in the study area. The decision as to the driving times to be used would depend on whether or not there are large differences between peak and off-peak driving times in any particular segment of the study area. For example, if the ratios of peak to off-peak driving times were about 2.0 for most segments of the transportation system, then this would mean that the relative driving times between the various zones would be about the same, regardless of which times were used. Under such circum- stances, it may be best to use off-peak times since about two-thirds of the daily travel occur during the off-peak period. On the other hand, if certain parts of the transportation system exhibit more congestion during the peak hour than others, then it would be more desirable to use some combination of the peak and off-peak driving times. Several studies have weighted traveltime in the ratio of two-thirds of the off-peak time to one -third of the peak time, since this is about the pro- portion of travel occuring during these two periods. 3« Trip purpose classification . —A decision must be made as to how many and what combinations of trip purpose categories will be considered in the study. Gravity model trip distributions have been made using as few as one trip purpose and as many as nine or more. There is no clear agree- ment on this point. The number of purposes is partially a function of II -3 the size of the urban area involved. In selecting the trip purpose classi- fication, it is desirable to take into consideration the volume of trips, the stability of these volumes, and the trip length characteristics for each trip purpose category obtained in the home interview survey. The amount of data preparation time, computer time, and analysis time, must also be con- sidered. It is important that the trip purpose classification system used be designed to sufficiently characterize different travel patterns in the area. On the other hand, it is also desirable that the trip purpose breakdown be realistic from the standpoint of forecasting. Most of the recent gravity model studies in smaller urban areas have used the following trip purposes and categories with satisfactory results: a. Home based work trips Those trips between a person's place of residence and his place of employment for the purpose of work b. Home based nonwork trips Those trips between a person's place of residence and all other places except his place of employment c. Nonhcme based trips Any trip which has neither origin nor destination at a person's place of residence, regardless of its purpose A study of travel patterns in Sioux Falls, South Dakota (k) , indicates that the differences in the accuracy obtained when using six trip purposes, as compared to using the above three purposes, are insignificant in this small area. The computer programs to be used with this manual accommodate only three trip purposes, although more purposes could possibly be used with manipulation of the programs. Generally, in small areas, a three purpose model should be sufficient. In some large areas where eight general purpose categories have been used, it has been observed that the results could have been improved with even further stratification of shopping trips to distinguish between convenience shopping trips (trips to grocery stores, etc.) and other shopping trips. h. Treatment of external trips. --Finally the treatment accorded to external trips, trips with one or both ends outside the cordon line, presents the transportation planner with two choices. First, the external cordon stations could be considered as fictitious zones and assumed to produce and attract trips in a manner similar to the internal zones. In other words, all trips would be treated alike. Generally, it is undesirable to do this for the following reasons: a. Trips made by those persons living inside the cordon may exhibit different trip length characteristics than those made by persons who live outside the area. II J+ b. External -to -external trips are associated with the study in question for only a small portion of their total journey and therefore exhibit distribution characteristics which have nothing at all to do with the study area. Second, the total universe of trips could be treated as three distinct types as follows: a. Trips made within the study area b. Trips made outside the study area but with one end of the trip inside the cordon c. Through trips with both ends outside the cordon For the first type of trips the gravity model can be used directly. For the second type of trips, the gravity model is also used. However, since the trip length characteristics of the second type of trips may be different from those of the first type, a separate gravity model analysis should be made. This decision is particularly important in the smaller urban areas where the proportion of these trips to the total trips in the area is quite large but the total number of trips is small. If the combination of type (a) and (b) trips were used to develop a single set of traveltime factors for a small urban area, the results may be distorted, sending the trips of the first type too far, and the trips of the second type not far enough. Consequently, it is desirable to treat the type (b) trips as a separate trip category and develop a gravity model by treating them as a single purpose . For the third type of trips, through trips, a growth factor technique such as the Fratar procedure can be used. II -5 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER III - DESCRIBING THE EXISTING SYSTEM AND ASSIGNING TRAFFIC September 1965 Once the trip purpose classifications to be used in the gravity model distribution have been determined, the next decision is the method to be used for assigning traffic. Traffic may be assigned by either an all-or-nothing procedure or by a diversion procedure. In order to assign the survey trips to the present network , the following operations must be accomplished: A. Preparation of the network B. Editing the network C. Building and checking the minimum path trees D. Processing the survey trip cards E. Assigning 0-D trips to the present network A. Preparation of the Network The travel facilities inventory provides the information which is used for defining the existing network to the computer. The information that is required for each link used in the traffic assignment network is: l) the link speed or traveltime and the link distance; and 2) the capacity of the link, if a capacity restraint procedure is to be used. There are several items involved in preparing the network and these will be discussed individually. References will indicate that a more detailed discussion of an item is available in another publication. 1. Map preparation . - -Two base maps of the study area are usually required for defining the street and highway system. Another map showing the traffic survey zones and boundaries and their respective zone numbers is used to locate the zone centroids. In addition, appropriately scaled street and highway maps must also be available. 2. Speed and traveltime information . --The speed and traveltime information is used in coding the links making up the transportation network. Normally, distance and speed are coded for each link and the computer calcu- lates the time. The time may also be coded, but this will be more difficult to adjust later, if adjustment is necessary. Reference (l), chapter III, page 20, provides more information on speed and traveltime. III-l 3. Traffic volumes . --Traf fie volumes should be obtained for as many- streets as necessary to describe the vehicular movements on the existing major street system. This information is not used in describing the network, but it permits the evaluation of the initial traffic assignments. h. Street capacities . --The capacity of each link in the coded street and highway network must be determined if capacity restraint procedures are to be used. The information from the travel facilities inventory is used in computing the link capacities. 5« The network description . --The traffic volume data, and the speed and traveltime for each link would be coded on the base maps. With this informa- tion now available, the network description process can be initiated, as follows: Locate and number the zones and centroids Define the basic transportation network Connect the centroids to the arterial street system Locate and define the nodes Prepare a list of nodes Assign the node numbers Set turn penalties and turn prohibitors Define the link parameters Locate and number the zones and centroids . — The establishment of sectors and traffic zones should consider the requirements of the traffic assignment procedure and the related computer programs as well as the require- ments for data collection. In addition, planning areas, census tracts, and the requirements with regard to traffic forecasting areas should also be recognized. The trip table builder and trip end summary programs require that the external sectors be numbered higher than the internal sectors. In a traffic assignment, all trips are assumed to be loaded on the highway network from a single point established for each zone. The point of loading for each zone, defined as a centroid or loading point, should be located at the center of activity for the zone. For a completely residential zone, the center of activity would be the center of gravity of the zone * s popula- tion. For example, consider the typical zone shown below. Assuming each spot represents 100 persons, the center of population, or centroid, would be established approximately as shown. For mixed land use zones, such as residential with commercial, the location of the centroid is determined to a large extent by judgment. The size of the zone must be determined by keeping in mind the statis- tical stability for sampling, the proper loading for assignment, and the limitations of the computer programs. On the copies of the zone maps that were previously prepared, the centroid of each zone is located and marked. There is one centroid for each survey zone and external station. They are numbered in a consecutive unbroken sequence beginning with the number 001. A transparent overlay is then placed on the street and highway maps and the centroids with their corre- sponding numbers are transferred to this overlay. b. Define the basic transportation network . --The selection of a network for assignment purposes is largely dependent upon knowledge of the area and judgment. The data required for selecting the network are the street classification map, traffic volumes, and street capacities. All streets that carry a substantial volume of traffic should be included. Naturally, a substantial volume means something different in each city. In a large city, it may mean 5,000 vehicles per day. In smaller cities, the number might be 1,000. As a general rule, all expressways and all arterials should be included, as well as a portion of the collector streets. The local streets are not included but are simulated by connections between zone centroids and arterials. The assignment procedure does not assign intrazonal trips since all trips are loaded to and from a single point, the zone centroid. Therefore, if all streets are included in the system, the assigned volume would tend to be lower than the actual volume counts. On the other hand, if too few streets are included in a network, they would tend to be overloaded; the above factors are compensating to some extent. In general, a good guide is to include all streets that are protected by a signal or a stop sign. When there is doubt about whether to include a particular facility, it is better to include it than to reject it. Each facility that is selected for use in the network is traced from the base map on the overlay that contains the centroid locations. Refer to the sample map in figure III-l, the sample network in figure III -2, and the sample base map in figure III -3- c. Connect the centroids . --Each loading point or centroid must be connected to the arterial street system. Because of computer program restrictions, a centroid can have no more than four connections to the system. As these are hypothetical links that represent the local street system, they are drawn as dashed lines at an angle to the arterial street. Centroids are not normally located directly on a link of the system. If they should fall on a link, they must be relocated adjacent to it and connected by a link of zero traveltime and distance. An option in the computer III -3 Figure III -4. Sample zone map. Ill -4 Figure III-2.-- Sample network map III -5 Figure UI-3- Sample street classification and "base map III -6 program allows trees to be built through the zone centroids. It is recommended that a centroid be given connections which are con- sistent with reality (with a maximum of k allowed). More connections tend to smooth the traffic on the adjacent links. If only one connection is given to the centroid, the point at which it connects to the arterial street system will show an abrupt change in traffic volumes. This should be avoided, but for those centroids that represent only a few trips, it may be sufficient to connect them with only one link. When in doubt about the number of connecting links, the larger number of connecting links should be used. The centroid connections are shown on the sample network map in figure III -2. d. Locate and define the nodes . --A circle or small dot is placed at each intersection in the system to denote the nodes. One of the limitations imposed by the computer program is that there may be no more than four outbound links from a node. When intersections of more than four outbound links are encountered, it is necessary to add extra nodes at the intersection in such a way that none of them has more than four outbound links. The following example shows how this may be done: DESIGN SIMULATION > — Zero time and distance link If a link is expected to have a traveltime greater than 9.9 minutes, a node or nodes must be inserted in the link so that no link will exceed this limitation. The overall limitations of the computer programs must also be recognized. These are: l) no more than 100 links are allowed in a minimum time path; 2) a maximum of 699 nodes is allowed in the total present and proposed system and 3) the total time in a minimum path must not exceed 99. k minutes. At this point, all one-way streets should be marked with arrows in the direction of travel and, if space is available, some of the major geo- graphical landmarks should be identified such as bridges, major streets, etc. If the system that is being coded contains some high type facilities, such as freeways, their interchanges may be coded directionally. In directional coding, turning movements and weaving movements may be specified as links in the system. Thus, the longer distance and traveltime in the loops of a cloverleaf interchange may be simulated in a network by directional coding. Some examples of directional coding are illustrated in figure III-^. Ill -7 Fwy. ,' Fvy. Figure III -k. — Examples of directional coding. e. Prepare a list of nodes. --Before numbering the nodes that have been previously identified, a list of node numbers is prepared as shown in table III-l. The IBM 1620 tree building and assignment programs require that the centroids be assigned the lowest of four ranges of numbers, beginning -with the number 001. The next highest range of node numbers is assigned to the external station nodes. The centroid and station node numbers should be consecutive. The third range of numbers denotes the arterial nodes and the last range of numbers is assigned to the freeway nodes. The highest code number that may be used is 699- There may not be gaps in numbering within a range of nodes. There should be a gap between N3 and N^. For proposed systems, a break should not occur between the ranges of arterial node numbers and freeway node numbers (Nr and Ng). The gaps are provided in this manner so that revisions and corrections to the network may be easily made without destroying the continuity of the III -8 numbering scheme. Proposed arterial street nodes would be added to the lower end of the arterial node range. Table III -1. --List of nodes Node group Zone centroids Station nodes Arterial nodes Freeway nodes where : N 2 = V N 3 N 7 s 200 » 699 N ? - \ s 1^0 Range of node numbers 1 to N N 2 to N 3 N^ to N 5 N^ to N ? (max. of 1*1-0, if used) Special note must be made concerning the numbering of the freeway nodes. If a diversion assignment is to be made, a maximum of lUo nodes may be used to describe the proposed street or freeway system. The lowest numbered of the possible 1^4-0 nodes must be of a higher number than the zone centroids, station locations, or arterial intersections. In this case, there may be a gap between the arterial and freeway node numbers. There should not be a gap between the arterial and freeway node numbers if an all-or-nothing assignment is to be made for an entire existing system with freeways or a system combining present and proposed freeways. The existing freeways should be coded with the lowest numbers in the freeway node range to facilitate the possible use of a diversion assignment to alternate proposed highway plans. A more detailed discussion of diversion assignment can be found in section E. of this chapter. If there is no diversion assignment to be made, the computer will make no distinction between arterial and freeway nodes. But, it is still recom- mended that a definite range of node numbers be assigned to the freeway system. When this is done, there cannot be a gap between the arterial node and freeway node numbers. When coding the ranges of nodes on the computer program control card, the freeway and arterial node numbers are combined into one range of numbers, without a break, and coded as the range of arterial nodes. With the assignment decision made, the "available node" table is now prepared. It shows all of the node numbers and the node groups to be used in the numbering system. When a number is assigned, it would be removed from the table to prevent the assignment of duplicate node numbers. After the node numbering is completed, the table would be retained to provide III -9 a ready source of available node numbers when making corrections or additions to the network. f. Assign the node number s .- -Many advantages accrue from adopting a systematic method of assigning node numbers. In general, it has been found best to proceed along the main radial highways, from the center of the urban area outward, and to complete the numbering in the sector between two radials before proceeding to the next. Then the nodes in the same numerical range are grouped together. This facilitates the plotting of trip volumes and other tabulated data during the analysis of the computer runs. The node numbers may be written either beside the dots representing the nodes or, if preferred, they may replace the dots. Legible writing is critical, as this is the master tracing that will be used throughout the traffic assignment. After completing the numbering, the maps should be reviewed to be sure that every node has been assigned a number. g. Set turn penalties and turn prohibitors . --In order to provide for turn prohibitors or to penalize certain turning movements, it will be necessary to use four nodes and dummy links. The diagram shown in figure III -5 shows an intersection which required turn prohibitors and turn penalties It should be noted that turn penalties and prohibitors will reduce the number of available nodes considerably, as they require four nodes per intersection. The following discussion will show the difference between the normal inter- section and the one which would require turn prohibitors or turn penalties. A. NORMAL t6 5 .7 FOR CODING WITH TURN PENALTIES OR TURN PROHIBITORS Figure III -5 — Turn penalties and turn prohibitors The following example illustrates the method of coding turn prohibitors and turn penalties for the normal intersection shown in figure III -5 A. To begin the process, the intersection node, 5, is eliminated and four nodes are added. Since each link is coded direct ionally, a turn prohibitor can be coded by just not coding the link in the direction to be eliminated. To establish a turn prohibitor for the movement from node h to node 3, this direction is not coded. The movement from node 3 "to node k is coded to allow the left turn from the other direction. If the links from node 8 to node 6 vere one-way, the movements from node 1 to node k, node 1 to node 2 and from node 2 to node 3 would also be prohibited by not coding them. A zero time would be coded for each dummy link used. This procedure may not be totally effective, particularly in coding freeway interchanges. See the caution in a following paragraph. For coding turn penalties, the same configuration of nodes, shown in figure III -5 B, would be used. The only change is that an amount of time is inserted for the dummy link representing the movement to be penalized. Where turn penalties are deemed necessary, the amount of time to be coded would vary with the city. A turn penalty of 0.2 minutes has been used in smaller cities and 0.3 minutes has been used in larger cities. A caution about using nodes with dummy links for turn prohibitors should be noted here. After test trees have been built, they should be checked to see that the turns have been penalized or prohibited properly. It is entirely possible for the path to use the other dummy links to double back on itself and in effect actually permit a turn at a prohibited location. An example best illustrates this possible problem. Figure III -5 B shows the right turn prohibited in the direction from node 9 "to node 8. It is possible for this movement to be duplicated by going from node k to node 2 to node 3- If this happens, it will be necessary to adjust the time on each affected dummy link. h. Define the link parameters . --At this point, the overlays of the network showing the links, nodes, centroids, street names, turn pro- hibitors, etc. are complete. All further work is done on transparent (reproducible) prints or opaque prints of these maps. On a print, each link must be defined by its parameters --link distance and speed or traveltime. Dummy links are marked with zero time unless turn penalties are used. Measuring tapes are prepared from tracing paper and marked off in hundredths of a mile for each map scale. The measuring tape is marked with the title and scale of the map with which it is to be used. As each link is measured, the distance is written along the link. Any links which obviously will exceed the 9-9 minute limit for each link must be divided into two or more smaller links by inserting additional nodes. 6. Special treatment for external stations . --If the preceding method was used in coding the network, the links connecting the centroids repre- senting the external stations will be treated by the computer as local streets. As local streets, the vehicle -miles and vehicle -hours of travel on those links will be summarized in the local street category. To avoid this situation, an arterial or freeway node may be placed adjacent to the traveltime and distance link. This has been the usual practice. Some transportation analysts, however, prefer to code a traveltime for this hypothetical link which represents an average traveltime to or from that external station and the outlying zones. This will not affect the tree routing, only the summaries of vehicle-miles and vehicle -hours. Other analysts prefer to treat the external stations by increasing the size of the study area. They remove the external station and extend the network through to the outlying zones, each of which contains a centroid or loading node. This, of course, requires special handling in the building of trip tables and in the preparation of the network. It does permit a "diversion" of the trips approaching the external station, as it gives them a choice of alternate routes to use in entering the study area. 7- Identifying and storing the maps and tabulations . — Each traffic assignment is given an identification number. All of the tracings and prints for a particular system or assignment are kept together in a suitable filing device. Copies of the following tabulations (usually the original copy) are bound in hard covers and preserved for reference: a. A tabulation of the link data cards used in building the network b. A tabulation of the link times as built by the computer c. The tabulation of the link volumes as assigned by the computer d. If printed, a tabulation of zone-to-zone trips as used by the computer e. Tabulations of selected trees, and other information 8. Link data card preparation. --This next step involves coding and key punching the link data cards. To prepare for coding, a transparent overlay is placed on the network map and, as each link is coded, it is marked on the overlay. This procedure prevents the coding of duplicate links and permits simultaneous coding by more than one person. The card type 12 format is required for coding the link data as input to the traffic assignment edit and tree building program, see chapter XI, assignment 1 program description. 9. Checking for errors . - -The coding of the network continues until all of the links have been coded and checked. Then the link data cards are keypunched and A node and B node card listings are prepared. The B node listing is prepared first by sorting the link data cards on the first B node field and listing them with tabulating equipment. The cards are then resorted on the A field and again a listing is prepared. These two listings provide a permanent record of the link data cards and permit a preliminary manual editing of the information. At this time it is advisable to determine that each centroid has at least one connection to the arterial street system; that there are no more than four outbound links from a node; and that there are no misplaced card punches or illegal codes. Usually a brief scanning of these listings vill indicate some errors that should be corrected before any further processing. If errors are found at this stage, those cards that are in error are removed from the original card deck and replaced by corrected cards. An example of a coded network and its accompanying list of nodes is shovn in table III -2. B. Editing the Network Before the network can be edited and sample trees built, the type 11 and lk control cards for the edit and tree builder (assignment l) program must be prepared (see chapter XI). The type 12 link data cards are used as input data to the program. Enough passes through the program must be made with the control card coded for building only one tree. A description of the control cards and the error messages is included with the program writeup for the assignment 1 program. The cards are examined for the following: 1. Card type numbers and card sequence are checked for validity. Only type 11, 12, and 1^ cards are accepted. 2. Each node number is checked and any node coded as zero will be rejected. 3. A message is typed out if more than 1^4-0 proposed freeway nodes are specified in card type 11. k. The low freeway node number is checked for being higher than the other types of nodes when a diversion assignment is indicated. 5. The town code in each card is checked for consistency with card type 11. 6. Each "A" node is checked for being within the specified range of nodes. 7. Each connection to every "A" node is checked and a message is typed out for each connecting node that is outside the specified ranges of nodes. 8. The computed time is checked for being greater than the 9«9 minute maximum link time. 9. The path being built is checked for exceeding the maximum of 99*^ minutes . Ill -13 Table III -2. --Link data (partial listing) and example list of nodes, Test Tovn #20 K o t) tJ ■p o o o Node group Zone centroids Station nodes Arterial nodes Freeway nodes Node numbers 001 to 006 100 to 103 199 to 238 None 12 12 20 20 212 213 1 207 210 30 25 024 006 211 214 22 23 010 010 215 229 25 25 006 010 228 28 010 12 12 20 20 2U 215 211 212 25 25 006 006 213 214 23 24 010 010 215 231 24 35 010 013 12 12 20 20 216 217 240 100 35 40 010 020 005 216 37 35 011 013 217 218 35 28 013 003 222 35 032 12 12 20 20 218 219 217 218 28 40 00 3 028 219 220 40 35 028 009 12 12 20 20 220 221 219 003 35 28 009 007 221 220 30 30 005 005 235 224 40 35 030 02 5 12 12 20 2 222 223 217 241 35 40 032 013 203 200 26 30 020 010 225 27 02 12 12 20 20 224 225 203 205 32 29 010 010 221 223 35 27 025 020 226 227 35 36 020 024 236 45 030 12 12 20 20 226 227 207 242 30 24 010 001 224 103 35 40 020 030 228 225 37 36 024 024 229 30 006 12 12 20 20 228 229 212 213 28 25 010 010 226 227 37 30 024 006 231 230 32 35 014 014 237 45 030 12 12 20 20 230 231 D06 215 11 35 017 013 199 228 35 32 028 014 229 233 35 36 014 022 238 40 037 12 12 20 20 \2Z2 2 33 199 234 35 35 014 011 243 231 4b 36 029 022 006 243 18 45 004 02 7 244 40 28 III-lU 10. An error message is typed if the table of nodes yet to be extended is full. This is an indication that there is an error in the system of nodes that permits looping. 11. The reversals of links and their traveltimes are checked, if desired, to see if the opposite direction has been coded and to see if the traveltimes are equal. An error message vill be typed out for one-way links as well as links vith errors. 12. An error message is typed out if it is impossible to build the selected tree path without using two freeway nodes. The portion of the path involved would be typed out. This test is made only when freeway nodes are specified in the program parameter card. 13- A message is typed out if there is a node within the specified range of nodes for which there are no extensions. Ik . A message is typed out if a node specified in the range of nodes is greater than 699- C. Building and Checking Minimum Time Path Trees After the edit phase has been completed without errors, the program parameter card is changed and selected trees are built using one or more type ik parameter cards . The same edit and tree building program builds trees and is used to punch any number of the minimum path trees. More than one type ik card can be used, but the range of nodes punched in each card must be a range of continuous node numbers. As a usual practice, about 6 to 10 trees are built. These sample trees are chosen at the option of the pro- gram user, but it is advisable to build at least one tree from a centroid in the central part of the study area and several trees scattered around the peripheral area. This would provide good coverage of the study area for detecting illogical routings caused by network coding errors not checked for in the edit phase. Some typical errors made in the network coding are shown in figure III -6. One of the most common is the so-called "jumper 1 ' which allows the path to jump intermediate links and will short circuit the minimum path. The type 17 existing tree path cards are punched for each pass through the program. This is the reason that only one tree is specified during the editing. To format the cards for tracing and plotting, they are simply listed. To facilitate checking, vertical lines are drawn on the listing to separate the various fields and the fields are marked. In order to trace an existing tree path, start at the destination node (which is a "B" node) and go continually from the "B" node to the back node until the home node ("A" node) is reached. A sample tree is listed in table III -3, and the following example exhibits how the tree is traced from centroid (home node) 1 to centroid ("B" node) 6. First "B" node 006 is located along with its back node, 232. Second "B" node 232 is located III -15 ->" — +*- A. Dead end link B. Link coded in the wrong direction -V si "^ C Missing links D. Jumper link Figure III -6. — Common coding errors. along with its back node, 199. Third, go to "B" node 199 and obtain its back node, 230. Fourth, go to "B" node 230 and find its back node, 229- Fifth, "B" node 229 is located and its back node, 227, obtained. Sixth, "B" node 227 is located with its back node 225. Seventh, locate "B" node 225 and its back node, 205 . Eighth, locate "B" node 205 and its back node, 20U. Next, locate "B" node 20k and it is found that its back node is the home node, 001. This indicates that the tree has been com- pletely traced. Building and tracing trees for the combined existing and proposed system is covered in a later chapter. The manual plotting of the selected trees would be accomplished at the same time as they are traced. Each tree is plotted on a separate trans- parent overlay that is placed over the base map. Each link, formed by a "B" node and its back node, that appears in the trace is drawn on the overlay. A sample plotted tree is shown on figure III -7 for a sample city. Isochronal lines can be drawn on the selected tree traces showing the III -16 ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooo oooo Accumulated driving time from "B" g g g jj g g g g g jjg g * £ ° node to "A" node g g § o g g g g o g g g g g g ooooooooooooooo oooooooo ooooooo "Ovt^O^NN^mo^'J-OfO.i- .l-tOOOOHNON^NH^m the "B" node f\j<\j iiutic w «• iiuu.c OOOOOOOOOOOOOOO OOOOOOOOOOOOOOO OOOOOOOOOOOOOOO Back node which forms the link with ^NOM^OMoomecvooriH „ „ NHmOOHHMONNmOfnN the B node fjrMfsirsij-o ii— it , 00—tr\j<\j(\jraooooo(\i D noac OOO«NO0JOtMOO>f-^(MOOOOf^N^vJ-in(\(N'0 OOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOO oooooooooooooooooo oooooooooooooooooo "■d" n0( i e mrsi^r\jmor-fMO-^rn(Mifso^m-^ooo ° " WVJ - C MOOMO^OHOHH-oorooor\j(\ir-.o.oir>3-c\l>r>a-rr)r0^C0<-iO^>}-rn>}-r"-v0 a noae ooooooooo-«-<--<-«-.-.-4 C\lrco^^-^coo oooooooooooooooooo oooooooooooooooooo oooooooooooooooooo oooooooooo o o o o o o o o 13 noae oooooooooooooooooo NOtNJfSIO(MMC\IOMf\JMMf\|C\INO'-l oooooooooooooooooo oooooooooooooooooo oooooooooooooooooo 'o^Nr^»d-LAvON-c6tro— 'rMfnvTir\«or- "A" node ooooooooooooo— i—ioo—i (N(NiC\J(\|{NJ(M(NtNt\l(\lfMrs)f\J(N|rgfNJ(MrNJ OOOOOOOOOOOOOOOOOO Citv code f\i *+> or 5- The format for each card type must be described for the computer on a parameter card. The card formats are described in the program descriptions. The card output of the trip table builder are the stacked volume cards (type 03) or trip tables. These trip tables, by purpose, will have to be combined before assigning the trips to the present network. The combined trip table is obtained with the add trip tables program. The operation of this program is covered in the program description. E. Assigning 0-D Trips to the Present Network After the 0-D trip tables for each purpose have been added together, the total purpose trip table is ready for the initial assignment to the present network using the assignment 2 program. At this point, there has been no adjustment of the network other than correction of coding errors, and the minimum path trees have been built on the basis of the traveltimes coded into the network. No attempt has been made to account for other conditions that may affect the route choice between two points, such as: congestion, ?>edestrian interference, pavement condition, etc. The initial assignment to the present network is usually done on an all-or-nothing basis. The output of the all-or-nothing assignment is the directional and nondirectional link volumes. Examples of an all- or-nothing assignment are shown in tables III -5 and III -6. Ill -27 o o o o o oooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo oooopo ooooqoopp cnjooo^'oo'h^ooooooo Directional roo^>orooorsjrooooooo ,7 . oxuuaj - ^-o-i>oojoo>j-moooooo link volume ppo'Ct^ppcoppppppp ooooooooooooooo o o o o o o OOP o o o o o o "B" node (k) oooooooooo oooc ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo o ooooopoo o o o o a o O FoTo'o^Foo op >o vO "F"^ p^pmppmpppproror-o Directional o— ior\ippr-ppppmr<*>cocr 1 -t«b- •.r«l,,»n« POP^PPCOPOOOOOh-fSJ link volume ooooooooooooo-to OOP OOOOOOPPPPOO „ „ , v f\l lA O HF f\f O —i rt (M OTCF" u\ ~* O CO B node (3) noho^ohhh-kmhi^^-i fVMNJ fSJJ\JfVJP_f\J (M<\|C\lt\jr\l(\jr\Jf\J O O 0"0 O O p~p~~o P CO o o o poooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo ooooooooooooooo o o c o_o o o o o o o o OOP O 0~0"0 OO»0O-hOOOv0nJ-v0 r>4««.« + 4^v,„i oooo^>o>omc\jooomoo uirectionax o—iooororovo^ooooo*^ 1 Ink vol unw» O^OOH^vf^ooooCNO j.inK voxume ooooo-'-oooooooo O O OOOOOOOQOOOOO "■a" ~~a~. /o\ SO vO lf\;>0 & CO -* O ~* -J- m >*■ r- & a noae \d) o o o o'o oo— •— t^-Hi-Hr^i-ir-* fM CM CM PJJSSLXS1 4-ir\mr-ooooooomo>oo inmmmpmoor-oppocop link volume rr>>ovo-4-p»}-ir\p>rpppppp ppppp^pppppoprsjp OO P PPPPPO.OQ P P P P „_,, , ,.x (M^^NfriNsOin^NO-HMiro B node 11; pppppppppp-4^^pp cm p c\j cm rvj p rvi rg m nnnnqh OOOOOOOOOOOOOOO OOOOOOOOOOOOOOO ooooooooooooooo „ A „ m ^ m io.Koo o> o h N t?TTin >o c- A node oooooooi-t— mcihhhh L£SL£SLfcl (NJ.f^J^.CV_O...CsLJN.fsL.J^,CsLfSLOj.^SUN O.O r^ OOOOOOOOOOOOOOOOCOOOO OOOOOOOOOOCOOOOOOOCOO OOOOOOQOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOCOOOO OOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOO O.OQ5.0.0 O O Q_Q O..O QOOOOQ OOO O- OmOOOOOMO^-OOOOOOOOOO OOf-OOOOOOOf-OOOOOOOOOO -^O0^OOOOOf»OOOOOOO-a-ir\>or^cocr*o ^t\iro-4-ir»vor-coc^o» OOOOOOOOOOO-H-Hr-tr-IOOf-tOOfSI Cs) IO C\J CD < < en CL a: q: o or o X UJ _l o X UJ > X (- UJ -J Q. CC H > < ^ cvj g cd to sdmi xuom "lvioi iN30U3d v-5 trip purpose and travel mode category. This information is also used as a check in the trial and adjustment procedure for determining the effect of traveltime on trip interchange. Next, the mean trip length for each general trip purpose and travel mode category is obtained by dividing the vehicle -minutes of travel by the total number of trips in each of the respective categories. These data are also used in the determination of traveltime factors (Fy) as will be discussed in following sections of this volume. 2. Selecting initial traveltime factors .— At the present time, a specific mathematical equation or function which can adequately express the effect of spatial separation on zonal trip interchange is not available. A single exponential function of traveltime has been proven to be inadequate. Other functions are currently being tried and evaluated but to date have not proven satisfactory. Consequently, it is necessary to go through a trial and adjustment procedure (or a calibration, as it is called) to arrive at a quantitative measure of this effect. This quantitative measure is called a traveltime factor and its use in the gravity model formula is shown in chapter IV of this volume. In the preceding discussions, all of the basic inputs to the gravity model formula, with the exception of traveltime factors (F.jj), have been developed. Zonal trip production (P^) and trip attraction (a.) values for each trip purpose and travel mode category have been obtained with the trip end summary program. In addition, minimum path traveltime s (t. .) between all zones have been developed with the update skim trees program. It is the purpose of this section to discuss how the traveltime factors are developed. The initial set of traveltime factors for each trip purpose and travel mode category can be determined in at least two ways. First, one can assume that each traveltime factor has a value of one, or in other words, that traveltime has no effect on trip interchange. It is known, however, that this is never the case. Second, a set of traveltime factors can be borrowed from some other city of similar size. The results of one of these methods for determining traveltime factors are used in the gravity model formula, together with the skim trees, intrazonal and terminal times, and the zonal trip production and trip attraction values. 3- Iteration procedure to get final traveltime factors . --With the initial set of traveltime factors determined for each trip purpose and travel mode category it is now possible to calculate trip interchanges using the gravity model formula. This is done using the program which is described in chapter XVIII of this volume. Input to the gravity model consists of the following items for each trip purpose and travel mode category: V-6 a. Zonal trip production and trip attraction values as obtained from the trip end summary program b. Minimum path driving times (skim trees) between all zones as obtained from the traffic assignment 1 program ( These are not the updated skim trees.) c. The intrazonal driving times and terminal times d. Initial traveltime factors for each one -minute increment of traveltime e. Zone-to-zone adjustment factors (Kjj) if not equal to 1.00 (These adjustment factors are discussed in a later section .) Output from this program consists of the following items: a. A table of zone -to -zone movements as estimated by the gravity model formula, if the option is specified (This table is a card type 03 that is in the same format as the trip table obtained from the survey data using the trip table builder program, except that the trips are in a nondirectional production to attraction format, ) b. Two sets of type 30 cards (The first set contains the gravity model attractions (type 30a). The second set of type 30 cards (type 30b) contains the adjusted gravity model attractions. The type 30b cards are used for balancing attractions in the final calibration of the model. The attractions are adjusted in the following manner: The attractions used as input data are squared and then they are divided by the output model attractions to obtain the adjusted attractions,) c. A trip length frequency distribution, by one -minute traveltime intervals, of trip interchanges estimated by the gravity model formula (These distributions contain intrazonal trips,) The trip length frequency distributions are the output most directly used at this point. In addition to these data, the vehicle hours of travel are also obtained. The percentage of trips, for each trip purpose and travel mode category are plotted, on rectangular coordinate paper, versus their respective traveltime increments. For convenience, these curves should be plotted on the same sheet as the trip length frequency distributions obtained from the 0-D travel pattern inventories. Figure V-2 shows the 0-D and the gravity model trip length frequency curves for a given set of information. Also shown are the vehicle hours of travel and the mean trip length. The above information is sufficient to begin the trial and adjustment pro- cedure for determining the "best" set of traveltime factors for each trip purpose and travel mode category. Comparisons between the actual and the estimated distributions indicate the degree to which the traveltime factors were correctly chosen. A comparison is made, for each trip purpose and travel mode category, between the actual and the estimated trip length frequency curves in regard to the V-7 2 to it! CM q CO 00 0> o 1 o ro - IO CM q IO 0> < < w a. o QC O X id _] o X Id > X H o z id Q. QC > < \l /f* /! 4 A 'J "^^ ^^s S» — >v* — • \ • — • — 10 20 TRAVELTIME, MINUTES Figure V-3. --Log-log plot and curve of best fit for traveltime versus traveltime factor. Comparisons between these expressions for different urban areas can then be made more meaningful. Once this line has been drawn, a curve similar to figure V-3 can be entered and a new set of traveltime factors selected. For example, the new travel - time factor for 9 minutes would be 75; for 10 minutes it would be 70; for 20 minutes it would be 32 and so forth. These new sets of traveltime factors are then used in a repeat run of the gravity model program. Zonal trip production (Pj^) and trip attraction (A.) values and the minimum path driving times (t^*) between all zones remain as previously described for use in the first run of this program. This second run results in a second estimate of trip interchanges, and a new trip length frequency curve for each trip purpose and travel mode category. The trip length frequency curves for the second run are then compared with those resulting from the origin -destination survey and if this comparison does not meet the criteria another new set of traveltime factors is developed as previously described. The gravity model program is then rerun and the process repeated until the criteria are met. Operationally this trial and adjustment procedure for determining traveltime factors should take about three runs, assuming that reasonable first approximations of the final traveltime factors were used. Examples of traveltime factors are shown in table V-2. The factors for work trips are the same as the example which was completed in table V-l. E. Phase Four - Balancing Attractions for the Final Calibration When the mean trip length and the trip length frequency curves meet the established criteria, the model attractions should be balanced. The reason for balancing attractions is obvious if the gravity model formula of chapter IV is reviewed. There is no assurance that the trips attracted to each zone by the model will equal those trips shown as attracted to each zone as summa- rized by the trip end summary program. It is desirable to have the attraction figures equal on a zonal basis before the zone -to -zone interchanges are analyzed further. Balancing attractions is accomplished by using the type 30b gravity model cards in place of the type 30 trip -end summary cards and rerunning the gravity model program. Operationally, it takes only one run to balance the attractions. However, the trip length frequency distribution and the mean trip length must still check according to the same criteria as was used in determining the final calibration. In some cases, it will be necessary to make another calibration in order to properly balance the model attractions . F. Phase Five - Adjustment Factors for the Gravity Model 1. Topographical barriers. — Some of the gravity model studies conducted to date have shown that topographical barriers, such as mountains, rivers, and large open spaces, may cause some bias in the gravity model trip interchange estimates. For example, a recent study in Washington, D.C., (12) indicated that the Potomac River had some influence on trip distribution patterns. A Table V-2. --Final traveltime factors, Sioux Falls, South Dakota, 1956 data Work 3 185 k 150 5 125 6 110 7 100 8 085 9 079 10 067 11 061 12 057 13 050 Ik 0^8 15 0^5 16 010 17 002 18 000 Trip Purpose Nonwork Nonhome based 220 210 160 120 130 100 090 080 085 070 070 060 060 055 050 6kk 039 038 035 032 027 030 025 026 021 023 016 014 000 005 000 000 v-13 study in New Orleans, Louisiana, (13) indicated similar findings. A study- in Hartford, Connecticut (lb), indicated that the toll bridges crossing the Connecticut River also affected travel patterns. The nature of the influence of such topographical barriers is not precisely known. All of the above-mentioned studies have analyzed the apparent reasons ■why these barriers have influenced travel patterns in their own unique situa- tion. In Washington, D.C., (12) it was attributed to the fact that off-peak hour traveltimes did not correctly indicate the amount of congestion which was present on those bridges crossing the Potomac River. An analysis of congestion patterns in the region indicated that there was more congestion in the area of the Potomac River, than anywhere else in the region. From this analysis, it was reasoned that a more realistic measure of the travel- time on these bridges was required. In the Hartford study (1*0 it was attributed to the fact that tolls are collected on several bridges crossing the Connecticut River. Since travel costs can also influence travel patterns, it was concluded that this cost barrier should be reflected by increased traveltimes on those bridges on which tolls were collected. In New Orleans (13), the Mississippi River separates several parishes. Many persons cross this river by ferry boat, as well as over the bridges. It was concluded that traveltimes on these bridges should be increased to account for the effect of the long traveltimes necessary to cross the river by ferry boat. In each of the above cases, it can be noted that the effects of topograph- ical barriers were incorporated in the gravity model by inserting time penalties (or impedances) on portions of the transportation network. In each case it appears that the need for these penalties is influenced by our present lack of knowledge as to a precise measure of spatial separation between zones. It is important to point out that when time penalties are imposed on portions of the transportation network, these penalties must be taken into account in the frequency distribution of trips observed during the survey of travel patterns. This can be done by first changing the network and obtaining new skim trees. Then, the trip interchanges, for each trip pur- pose and travel mode, are reprocessed to obtain a new trip length frequency distribution. The revised trip length frequency distributions and mean trip lengths are then used as a base against which any subsequent gravity model calibrations (also reflecting the time impedance) are compared. 2. Developing zone -to -zone adjustment factors . — It has been pointed out that there can be factors, other than the initial input to the model which may affect the patterns of urban travel. Travel patterns may also be influenced by various social and economic conditions which, to date, have not been completely identified or quantified. The effect of these factors may be accounted for in the gravity model formula by the use of zone -to -zone adjustment factors (K-y). V-lk In the larger urban areas vhere there are many types of major employment and shopping, and recreation centers of different types, these adjustments may be significant. However, in the smaller urban areas, one will usually find it unnecessary to use Kj^ factors. Because of the limited research on this particular point, the underlying reasons behind the ICj, factors are not completely understood. However, several studies!/ have indicated that the following may influence our ability to identify the real causes for the need to incorporate zone-to- zone adjustment factors into the gravity model formula: a. The trip purpose stratifications used today may not be precise enough to account for all of the basic differences in travel patterns. Home based work trips provide an example. Most studies deal with these trips in total as one trip purpose category (i.e., total work trips). This has an important effect on the distribution of trips because total work trips cannot be distributed according to the type of employment opportunities that may be available in the urban area. For example, assume that all work trips produced by a particular zone are made by industrial workers. When distributing these trips with the gravity model, or any other travel model, the largest proportion of these trips would be sent to the closest zones with large employment centers, regardless of the type of employment that is available. This means that many of these industrial workers may be sent to large offices and commercial establishments mainly because of their proximity to the residences of these workers. However, if total work trips were to be further stratified according to white collar and blue collar workers this deficiency might be substantially corrected. However, such a stratification may create problems in forecasting, since it would be more difficult to forecast the place of residence and the employment opportunities for blue collar and white collar workers than it would be to forecast these variables for all workers. Similar problems may also exist in the other trip purpose categories because of insufficient stratification. Suspected items of this nature complicate the problem of isolating the deviations that can be legitimately attributed to phenomena associated with a Kj.j factor. b. In present practice, a set of traveltime factors is developed for each trip purpose and travel mode category. Since trips between all zones are used in developing these factors, they represent the average areawide effect of traveltime on trip interchange. However, there is some evidence which tends to show that traveltime factors vary by zone depending on the characteristics of the people who live in each zone and on the distribution of land use activities immediately adjacent to these zones. c. There is some evidence that factors such as income and occu- pation may influence the need for zone-to-zone adjustment. In Washington, D.C. for example, it was observed that low income workers were not as likely to be employed in the central business district as were higher income workers. This observation was made by direct comparison of zonal interchanges estimated by the gravity model formula and those from the origin -destination survey. 1/ See, for example, references (12), (10) and (lb). 7-15 Regardless of the reason for zone -to -zone adjustment factors, the need may possibly exist for incorporating them into the gravity model formula. Even with these many limitations on the understanding of the Kjj factor, tests to determine the extent of the required adjustments and procedures for incorporating .them into the model have been devised and used in several studies.?/ These procedures require an analysis of the differences between the trip interchanges calculated in the final run of the gravity model program and those measured in the travel survey. Limited experience has shown that it is for the large traffic generators that the gravity model trip interchanges must be adjusted. For example work trips between all zones and the central business district of an urban area may require adjustment. Occasionally trips between one city of an urban complex and another city within the same complex might require adjust- ment. Consequently, the procedure used to develop adjustment factors is to compare the trip interchanges between large traffic generators as estimated by the gravity model with those developed from an origin -destination survey. These comparisons are usually made graphically for work trips and then for the other trip purposes, if necessary. This is usually accomplished by plotting the pertinent movements. A district map is used in this analysis. Sector lines are drawn to denote major traffic drainage areas. A fictitious radial transportation route is assumed to serve each of these sectors. The movements between each district (by major trip purpose category) and the central business district are then manually "assigned" to the radial facility within the sector in ubich the district is located. These volumes are then accumulated to the central business district. Using this procedure, systematic differences which reflect the need for Ik ^ factors can be located. Further information can be found in reference (2). Once the zone-to-zone adjustment factors (%«) have been determined, it is necessary to incorporate these factors into the gravity model calculations. To do this, the gravity model program is rerun using the same input data except that zone-to-zone adjustment factors are used. The trip length frequency of this revised trip distribution pattern must then be checked against the origin-destination survey distribution to verify its correctness. In most cases, there will be no need to make further adjustments. 2/ See, for example, references (12), (13), and (ik). V-16 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS ASSIGNING AND TESTING THE GRAVITY MODEL TRIP DISTRIBUTION September 1965 A. General This chapter is concerned with developing final measures of the overall accuracy of the gravity model distribution by assigning the model trips to the transportation network. The results of statistical tests developed with the trip comparison program and graphical tests made manually are used in this analysis. B. Assigning the Gravity Model Trips to the Present Network After the calibrated gravity model trip tables (type 03 cards) have been obtained, they are assigned to the present network in the same manner as the 0-D trip tables (see chapter III). The nondirectional assignment (type 05 cards for the gravity model trips) can be compared directly with the 0-D type 05 cards from the assignment 2 program. The two type 05 card decks are used as input to the trip comparison program. C. Statistical Tests The statistical tests are generally applied only after the gravity model has been calibrated to a trip distribution pattern obtained from an origin -destination survey. The procedure is to compare the final gravity model trip interchanges with the trip interchanges derived from the origin- destination survey for each of the major trip purpose and travel mode categories. The trip comparison program is used to do this. Specific details on this program can be found in the program description. This program accepts the two type 05 card decks of trip information as input, one containing the survey, and the other the corresponding gravity model estimates. These data will be the nondirectional link volumes obtained by assigning the final zonal trip tables obtained from the trip table builder and gravity model programs. The trip comparison program initially produces a table of differences between the gravity model link volumes and those from the origin -destination survey as shown in table VT-1. Next, it separates the movements into a maximum of 15 categories depending on the survey volume. For each of these 14 -26 67 -11 5 5 30 20 72 -8 Table VI -1. --Trip comparison program output, table 1 SIOUX FALLS, S.D. WORK TRIPS (LINK VOLUMES) O.D. VS G.M. TABLE 1 - COMPARISON OF LINK VOLUMES ANODE BNCDE O.D. VOLUME G.M. VOLUME DIFFERENCE 10 80 10 20 29 20 92 80 85 5 40 15 -25 7 7 40 31 -9 110 78 20 1 264 1 2 66 1 267 1 277 2 231 2 237 2 238 2 2 50 3 228 3 235 3 236 3 2 47 4 169 4 198 4 199 4 206 5 225 5 233 5 2 34 5 244 6 275 6 284 6 2 86 6 288 7 311 7 313 7 314 8 289 9 192 9 200 10 170 10 180 10 181 11 182 12 159 12 175 13 138 13 140 13 141 13 158 7 7 46 -34 40 40 77 -15 33 33 61 -49 78 56 -12 6 6 10 -10 15 volume groups, the differences between each of the origin -destination survey and the gravity model movements may be tabulated according to magnitude into 15 difference groups. For each volume group the sum of the squares of the differences, the mean difference, the root -me an -square error, the number of 0-D survey trips, the percent root -mean -square error, and the number of gravity model trips are given. This information is punched out in a form ready for listing as shown in table VT-2. Infor- mation such as that shown in table VI -2 is used in the statistical analysis while that shown in table VI -1 is generally used for analytical purposes. Table VI -2 shows the frequency of occurrence of various levels of differ- ences between the gravity model and the origin -destination survey movements within the 350-399 volume group. It also shows the sum of these differences, the sum of their squares, the mean difference, the root -mean -square error with its percent, and the total trips from both sources, within this volume group. It can be seen that a total of six movements falls into this cate- gory. The most serious deviation between the gravity model and the origin - destination survey movements is minus 75* The remaining information is easily understood. The RMS error has associated with it the two -thirds confidence level. This means that for this volume group it may be stated that two -thirds of the time the difference between the origin -destination survey and the gravity model movements is between + ^3 trips. The relative error for the volume group is approximately +11.5 percent. This relative error, however, assumes that there is no error in the information against which the gravity model results are being compared. Of course, this is not true, since all sampling procedures* contain errors due both to reporting problems and to sampling techniques. Consequently, the previously calcu- lated relative error does not mean that the gravity model error is +11-5 percent; as the model was compared with the origin -destination survey results, which, being the product of a sample survey, have an inherent error related to sample size. The errors related to sample size can be estimated using the results of some past work by the Bureau of Public Roads. Figure VI -1 shows a series of curves which illustrate the approximate RMS error which can be associ- ated with different volume groups for different sample sizes. For the data shown in table VI -2, the origin -destination survey sample size was 12.5 percent. From the curves in figure VI -1, it can be seen that the percent RMS error associated with this sample size for a mean volume of 375 trips is about 27. percent. This means that for volumes of 375 trips, a 12-5 percent origin -destination survey sample will give an estimate of total trips which 2/3 of time is within + 27 percent of the value that would have been obtained from a 100 percent sample, if such a sample were possible. D. Analytical Tests Generally, the analytical tests to be discussed below are applied only when the gravity model has been calibrated using a trip distribution pattern VI -3 Table VI -2. --Trip comparison program output, table 2 TABLE 2 - FREQUENCY DISTRIBUTION AND ANALYSIS OF DIFFERENCES VOLUME GROUP 350 TO 399 DIFFERENCE FREQUENCY SUM OF DIFF. -70 TC -79 1 75 -50 TC -59 1 57 -30 TC -39 1 30 -20 TC -29 1 28 -10 TC -19 1 19 10 TC ALS 19 1 15 TOT 6 -194 SUM OF SQUARES 11144. MEAN DI RMS ERROR 43. .09 PER CEN TOTAL O.D. TRIPS 2240. TOTAL G -32 11.54 2046. VI -k 200 100 * "^ * •^ *«N 60 ^ *^ X *N ^ "^^ ^ v. Y. 60 ^ V k x o .«••«) (J7.D.U.S.) **- "*" WHERE ; •/. D.U.S.' DWELLING UNIT 40 K V ^>> IN 'ERCENT E v> 20 o io ^ n. ^%*, >ft r ^\ . ^C*., "^ £'u ^*\ \^" *S ^ *" ^n 1 "" U > J^v ^ 5 j \ > o < s « 2 < > < < < > > 5 3 3 O O O o < o < 2 I > > 3 3 3 3 3 3 O O o o o o Figure VI -1. Relation of percent root -me an -square error and volume for various dwelling -unit sample rates. VT-5 obtained from an origin -destination survey. The analytical tests need little explanation because the procedure for conducting them is almost identical to the method used to locate the need for zone-to-zone adjust- ment factors (Kj*). This has previously been discussed in chapter V. However, it is important to discuss three points in respect to the use of analytical tests. First, such tests are very necessary. The statis- tical tests give an excellent picture of the overall accuracy of the gravity model in reproducing travel, but they do not point out geograph- ical bias. The sector type of analysis described in chapter V, therefore, should always be made in addition to the statistical tests, particularly, since each sector, or traffic drainage corridor, is generally served by at least one major radial route, and any geographically biased error could have serious design consequences. Secondly, the use of the results of these analytical tests are invaluable in explaining the accuracy of the gravity model to policy makers or citizen groups who may be interested in the forecasting techniques used for arriving at transportation facility proposals. Statistical tests have only minor application when explaining the forecasting techniques to such groups. Finally, the output of the trip comparison program can be used for a graphical analysis. The differences between the gravity model and the origin -destination survey link volumes are plotted. The plot will give an indication of the error in the nondirectional assignment. This output cannot be used as source data for design purposes because the movements are between production and attraction zones instead of between origin and destination zones. The conversion of these movements to origin - destination movements is covered in the following section. E. Converting the Gravity Model Results to Directional Movements Once the estimated gravity model trip interchanges have been shown to reproduce the travel patterns surveyed by the field inventories within an acceptable degree of accuracy, it is desirable to convert the gravity model results to directional movements between origin and destination zones. The gravity model results yield nondirectional movements between every zone of production and all zones of attraction, which when assigned to the transpor- tation network produce nondirectional volumes. It is sometimes desirable to know directional volumes, but no provision has been made to convert the gravity model results to direct origin -destination movements using the IBM 1620 computer. At the present time, any trip table conversion must be done manually and will require considerable time to complete an average size system. Consider the following example. Zone 1 produces 100 work trips daily which are attracted to zone 2. This is shown schematically in figure VI -2. Each of these trips leaves home in the morning and goes to the zone of attraction and returns directly home in the evening. VI -6 100 «> **»» 4 A^cLn zone Figure VI -2. — Schematic gravity model trip interchange - case I. The gravity model trip interchanges, converted to directional origin - destination movements could then be shown as in figure VT-3. 50 work trips J nJSSnStton / Zone 2 50 work trips [ Origin zone Figure VI -3. — Schematic origin to destination interchange - case I. This is a rather simple case, hut this is essentially how the trip con- version is done. If, however, not all of the workers returned directly home in the evening, but, instead, one -half of them stopped to shop on the way home, then only 75 work trips would have been produced by zone 1 and attracted to zone 2 as shown in figure VI -h.±/ Of these, 50 or two -thirds of them, had the home as the origin zone and one third, or 25, originated at the work place. 75 work trips / Zone 2 *i Attraction zone Figure VI -k. — Schematic gravity model trip interchange - case II. The trip conversion process would then provide the movements as shown in figure VI -5. l/ The remaining 25 trips would be shopping trips, produced by zone 2. VI -7 Zone 1 \ 50 work trips J Zone 2 ^ as Origin J ^ as Destination I Zftne X \ £>5 vork trips / Zone 2 as Destination H I as Origin Figure VI -5. — Schematic origin to destination interchange - case II. The trip conversion process uses trip interchanges in which trips are represented between production and attraction zones. The percentage of total trip production which are origins and the percentage which are destinations must be obtained for each trip purpose and travel mode category. In case I, above, 50 percent of the productions were origins and 50 percent were destinations. In case II, 67 percent of the pro- ductions were origins and 33 percent were destinations. These percent- ages must be either assumed or preferably arrived at by an analysis of the origin -destination survey data. This analysis is rather simple in theory but it requires a considerable amount of time. Essentiall y all of the trip cards for each trip purpose or travel mode category are sorted to determine how many of them are produced by all zones of origin and how many are produced by all zones of destination. The required percentages are then calculated directly. These percentages, which represent areawide averages, are then applied to each zone . Since for nonhome based trips, the zones of production and attraction are the same as the zones of origin and destination, respectively, the percent- age split is always 50-50 for this purpose. Past experience has shown that a 50-50 percentage split will usually give good directional movements for the home based trip purposes also. The outputs from this process are directional movements between all origin and destination zdnes. This output can then be coded and assigned to the transportation network to obtain directional link volumes. VI -8 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER VII - PREPARING THE PROPOSED STREET AMD HIGHWAY NETWORK September 1965 A. Coding the Network The proposed street and highway nodes are coded into the netvork in the same manner as the existing system nodes. Chapter III describes the preparation of the existing network and allowance is made for the addition of a proposed system range of nodes. In coding a proposed network for a diversion assignment, there are several items which should be noted. First, when coding future freeway nodes, the decision to make a diversion assignment must have been made during the coding of the existing network. If a diversion assignment is to be made, only the freeway nodes will have type 06, 07, and 08 cards punched for them. To obtain type 06 or 07 turning volume cards for any arterial inter- section node, the node must be coded so that it is in the freeway node range but it must not be connected to any other node in the freeway node range. Second, when adding proposed arterial streets to the network, these street nodes should be added at the bottom of the arterial node range, otherwise the traffic will be assigned to the proposed arterial streets by the diversion curve, instead of by an all-or-nothing assignment. The example list of nodes for the existing and proposed system in table VII -1 can be compared with the existing system example list of node6 in table III -2. Table VII -1. — Example list of nodes Node group Node numbers Zone centroids 001 to 006 Station nodes 100 to 103 Arterial nodes 199 to 238 Proposed freeway 239 to 2kk If an all-or-nothing assignment is made for a proposed network, only the freeway nodes will have type 06 and 07 turn volume cards punched for them. In order to obtain turning volumes for any other intersection, the node must be coded as was described above. There are no type 08 cards, diversion phase data, punched for this type of assignment. VII -1 B. Editing the Proposed System Network Described in chapter III, section B, are the details of editing a network, except that the passes through the edit and tree builder must now be made in accordance with the instructions for editing a proposed network. These instructions are included in the program description for the assignment 1 program. The program control card for the edit passes through the program should be coded for punching only one tree while editing the proposed network. Punching only one tree will expedite the editing. C. Building and Checking Proposed System Trees After editing, the program control card is changed to build selected proposed system trees. The same general procedures that are described in chapter III are followed except that the assignment 1 program punches three types of cards (17, 18, and 19) in producing a proposed system tree. Instructions for the operation of the assignment 1 program are included in the program description. In order to trace a proposed system tree path, the starting point is the destination node (which is a "B" node) in the type 19 card. The path is followed continually from each "B" node to its "back" node until a "back" node of zero, 000, is reached. The zero "back" node indicates that it is necessary to switch to the type 18 cards for this same home node number. The "B" node which had the zero "back" node is located as a "B" node on the type 18 cards. This indicates the node of the intermediate link, and the "back" node shown is used as the connecting node for the type 17 cards. The path is then traced from this "B" node on the type 17 cards. When the "back" node that is equal to the home node is reached, the proposed system tree has been traced. An example of this procedure is followed through for one tree in table VII -2. Sample proposed system minimum path trees are listed in table VII -3. Using the example trees for centroid ("A" node) 1 in table VII -3, one can follow the path from centroid 1 to centroid ("B" node) 6. This is the example shown in table VII -2. Table VII -2 . - -Tracing a proposed system tree Step no. "B" node 1 006 2 232 3 2U3 k 2^3 5 242 6 227 7 225 8 205 9 20^ Back node Card type used 232 19 2^3 19 000 19 2k2 18 227 17 225 17 205 17 20U 17 001 ("A" node) 17 VII -2 Accumulated driving time, "B" to "A" node Back node for link with the "B" node "B" node Accumulated driving time, "B" to "A n node Back node for link with the "B" aoife Accumulated driving time, "B" to "A" node Back node for link with the "B" node oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo oooooooooooooopooooooooooooooo *oort-HSN'HtOrJir\o-oo>j , (>vO(*iiAcoinoKio>uMna3^ >Orom<\ioOvOOiTi<^h-fOrr>^r»-%j-0«4-fO(Mifv>0*'-^^«NiO _£f. c\iOO^.^^^O^< \^*■ 0~>i- HO>NNHMOO>-t^OffHNninOCl ^•IOOOOHNOfgO l NH(<\^On>tNOOH^HNH^(nO>t „£JJN. <\J O oo-HC^o , >o-^«oor , -r>-coo N 0'OOi?\h-r^oo}-ir\«oo , o^ir\f^>o u^HH^Hlr^co^o^N(^^(^^ooolnc^m'0^oml-lo«t^t^u^^ ^tMCjr>^-l^-l^l^-iri^r-<^-lt>tM^Qfvl^(\ifMI\JfSlrufSifVJ^(MfSi^^H OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOOOOOOOOO P P P Q O Q ..P O P. P. P O.P P. PPPPPPPPPOPPPPOO HNONmONOOIftO'OOKllMOHNONNO'fOHNflOOO ■< «4- J- c-v r\j sj- o ooo^oo^HHNM^mm^^ooooooHH^NNinmf Q^^N.tNifsifsirjfNifsirjrjfsipjfsiQQririfsirsifsiNfsiNrjrjfNrjfN oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo Accumulated driving time, "B" to "A" node Back node for link with the "B" node "B" node Home node ("A" node) City code Card type oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo OOOQOOQOOOOPOOOOOOOOOOOOOOOOOO cof\ir-i-if\jif\rr»f^rOvj-(^fr»r«-r^^-ipoeo(Mf^i-imf\jf^r>-h- oo O m ^ O OnNOOOH^ONNmnN0(^fvJ>00t(D(NJO^N'OW(M>0000qooooooooooooooooooooooo oo "d> 6 J In vO -^ 5 < i j-r^o«t^OCT>fnfM— ioirifor-oo^>-<(vjo>o ioincioo>tcot , ici^o v >} , r > -(7'0'-iir\r-i('\co»o>l'cO'0'-icoa3 oooooooooooooooooooooooooooooo oooooooooooooooooooooooooooooo -PO.PQQOOQPQQOOQQQQOQQQPQQQPQQQQ 0'0^-i'J , >OOM>J , inNr4J3nOMOOHOK>^0>OONNncOO O^^OOOHCJMONPimN^^O-H^ON^^^NNN^NO Qosi(NjOf>joj(Njf\i(>jr\jfNjrj(Ni(Nifsi(\iQojfNi(NiruiNc\ifNifsir>jfvicNjrvjo ooooooHHMMNnn, and 16, only the data involved are typed. A correction can be typed into the INPUT area (for all except ER's 13, Ik, and 16), if feasible, and branching to location 00738 will continue the program. If the error cannot be corrected by typing into the input area, run out the cards in the hopper, correct the cards, load all the data cards back into the read hopper, and branch to the start of the program, location 00^02. Typed Message ER 1 Machine Action Halts after typing input area. Explanation Incorrect card type, only types " 11, 12, and Ik accepted. Correct if feasible, see "NOTE" above or if a blank card, branch to location 00714. Halts after typing input area. One. but not both, of the nodes in the range of node types is equal to ZERO. See "NOTE" above. ER 3 Halts after typing input area. More than 140 proposed street and freeway nodes specified in card type 11, see "NOTE" above. Halts after typing input area. The low freeway node is not of a higher number than other type nodes and a proposed system is indicated,, See "NOTE" above. ER 5 Halts after typing input area. City code in input card is not the same as the code in card type 11. This is checked for consistency by all programs. See "NOTE" above. ER 6 ER 7 Halts after typing input area. Halts after typing input area. "A" node is not within the range of nodes. See "NOTE" above. An extension of an "A" node is not within the range of nodes. See "NOTE" above. XI-4 Machine Action Explanation Halts after typing input area. The computed time for a link is more than 9»9 minutes. The link referred to is the first one in the type out without a driving time. See "NOTE" above for correction and restart procedures. Halts after typing input area. Minimum time path being built has developed a time greater than 99*^ minutes. To obtain the portion of the path which has been built, branch to location 66ll4 (A50). Halts after typing input area. Table C is full. This table contains the nodes yet to be extended. This is an indication that an error in the system of nodes is allowing looping. See "NOTE" above. Types out a link and a time for which an opposite and equal cannot be found. Continues processing. Types out the node for which there is no existing system node for continuing the the tree. Continues processing. Opposite extension or time for the link are not equal. Possibly a one-way link. Computer checks the entire system for this condition without stopping if Program Switch 2 is off. This error indicates the card type 12 which may need correcting. The entire system needs reprocessing unless these are one-way links. Impossible to build an existing route minimum path tree without having to use two freeway nodes in succession. Correct system and reprocess the deck of type 12 cards. Types out the node in error. Continues processing . Indicates there is a node within the specified range of nodes for which there are no extensions. XI-5 Typed Message ER 17 ER 18 H. TIMING Machine Action Explanation Halts after typing A node in the range of nodes the range of nodes is greater than 699. The from the card type 11. type out is the range of nodes. See "NOTE" above. Halts after typing the card type 11 which is supposed to contain the map scale factor. No map scale conversion factor in the card type 11, and the link time needs to be computed for a link. See "NOTE" above. A 366 node system required 17 minutes for processing and punching three sample trees (card type 17) and the link times (card type 02). There were no errors printed out in this run. A separate run to obtain skim trees for the same 366 node system required 15 minutes for processing and punching skim trees for 84 zones. I. METHOD The program first reads the type 11 card which establishes the table limits. Next, the program reads the type 12 cards and edits them for applicable errors. If program switch 2 is off, each link is compared with its reverse and error 13 is printed for each link which is not equal to its reverse. The type 14 card is then read and the trees are alternately built and punched for each node specified. After punching the last tree, the program checks for another type 14 card. If there are no additional type 14 cards, the type 02 cards (links and times) are punched and the end-of-job message is typed. If there are additional type 14 cards, the specified trees are built and punched and then, the type 02 cards are punched. XI-6 Card Columns 1- 2 3- 5 15- 17 The minimum time path trees in this program are calculated using the procedure developed by Edward F. Moore of Bell Telephone Laboratories in 1957* This algorithm finds the shortest path through a maze. In general, this shortest or minimum pat-n procedure consists of accumulating and saving the minimum time and path from a starting point, generally an origin zone, to an ever increasing number of points surrounding this point until all destinations are reached. J. INPUT CARD TYPES - Card columns not specified must be blank. 1. Card Type 11 - Parameter Card Content Card type, "11" City code Lowest number in range of zone centroid nodes, should be 001 18-20 Highest number in range of zone centroid nodes 21-23 Lowest number in range of station nodes 2^-26 Highest number in range of station nodes 27-29 Lowest number in range of arterial nodes 30-32 Highest number in range of arterial nodes 33-35 Lowest number in range of freeway nodes (when used) 36-38 Highest number in range of freeway nodes (when used) 55-59 Distance (map) scale factor: 1 mile = XX. XXX map inches or 1 mile = 01.000 miles XI-7 2. Card Type 12 - Link Data Card The program checks to see if driving time is specified for each link. If the driving time is specified, the program uses it and continues to the next link whether travel speed and distance are specified or not. If driving time is not specified, it is computed using the travel speed and distance. Therefore, a combination of both cases can be handled in the same type 12 card, if desired. Less than four extensions of a node are handled automatically. Card Columns Content 1- 2 Card type, "12" 3- 5 City code 6- 8 "A" Node 9-11 First "B" node 13— 1^- Travel speed in miles per hour (XX ). 18-20 Map inches or ground miles (XX.X). 25-26 Driving time (X.X). 27-29 Second "B" node 31-32 Travel speed in miles per hour (XX ). 36-38 Map inches or ground miles (XX.X). 43-44 Driving time (X.X). 45-4-7 Third "B" node 49-50 Travel speed in miles per hour (XX). 54-56 Map inches or ground miles (XX.X). 61-62 Driving time (X.X). 63-65 Fourth "B" node XI-8 Card Columns Content 67-68 Travel speed in miles per hour (XX). 72-74 Map inches or ground miles (XX.X). 79-80 Driving thne (X.X). See E - Options and Card Co n umns 1- 2 3- 5 Card Type 14 - Control Card There may be more than one card type 14. Switch Settings. Content Card type, "14" City code Low node number in the range of trees which are to be built and punched. This would be the centroid number when obtaining skim trees. High node number in the range of trees which are to be built and punched. If only one tree is to be built and punched, code high node equal to low node. This would be the centroid number when obtaining skim trees. K. OUTPUT CARD TYPES - Card columns not specified may be zero or blank. 1. Card type 17 - Existing System Trees Card Columns Content 1- 2 3- 5 9-11 9-11 12-14 Card type, "17" City code Home node ("A" node of origin for this tree) "B" node of destination Back node which forms the link of the path with the "B" node in columns 9-11 Card Columns Content 18-20 Accumulated driving time to the "B" node in columns 9-H from the "A" node (XX. X minutes). 27-29 "B" node of destination 30-32 Back node which forms the link of the path with the "B" node in columns 27-29 36-38 Accumulated driving time to the "B" node in columns 27-29 from the "A" node (XX. X minutes). ij-5-^7 "B" node of destination ^8-50 Back node which forms the link of the path with the "B" node in columns ^5-^7 5^4-56 Accumulated driving time to the "B" node in columns ^5-^7 from the "A" node (XX. X minutes) 63-65 "B" node of destination 66-68 Back node which forms the link of the path with the "B" node in columns 63-65 72-7^ Accumulated driving time to the "B" node node in columns 63-65 from the "A" node (XX. X minutes). To trace the existing system path from a home node ("A'* node) to any destination ("B" node), it is necessary to begin at the destination node and use back nodes to reach the home node. The path is traced backwards. 2. Card Type 18 - Intermediate links between present and proposed system trees Card Columns 1- 2 3- 5 6- 8 Card type, "18" City code Home node ("A" node of origin for this tree) Card Columns Content 9-11 "B" node of destination 12-14 Back node which forms the link of the path with the "B" node in columns 9-H 18-20 Accumulated driving time to the "B" node in columns 9-11 from the "A" node (XX. X minutes). 27-29 "B" node of destination 30-32 Back node which forms the link of the path with the "B" node in columns 27-29 36-38 Accumulated driving time to the "B" node in columns 27-29 from the "A" node (XX. X minutes). 45-47 "B" node of destination 48-50 Back node which forms the link of the path with the "B" node in columns 45-47 54-56 Accumulated driving time to the "B" node in columns 45-47 from the "A" node (XX. X minutes). 63-65 "B" node of destination 66-68 Back node which forms the link of the path with the "B" node in columns 63-65 72-74 Accumulated driving time to the "B" node in columns 63-65 from the "A" node (XX. X minutes). 3. Card Type 19 - Proposed System Trees Card Columns Content 1- 2 Card type, "19" 3- 5 City code 6- 8 Home node ("A" node of origin for this tree) Card Columns Conten t 9-11 "B" node of destination 12-14 Back node which forms the link of the path with the "B" node in columns 9-11 18-20 Accumulated driving time to the "B" node in columns 9-H from the "A" node (XX. X minutes). 27-29 "B" node of destination 30-32 Back node which forms the link of the path with the "B" node in columns 27-29 36-38 Accumulated driving time to the "B" node in columns 27-29 from the "A" node (XX. X minutes) 45-47 "B M node of destination 48-50 Back node which forms the link of the path with the M B" node in columns 45-47 54-56 Accumulated driving time to the "B" node in columns 45-47 from the "A" node (XX. X minutes). 63-65 "B" node of destination 66-68 Back node which forms the link of the path with the W B" node in columns 63-65 72-74 Accumulated driving time to the "B" node in columns 63-65 from the "A" node (XX. X minutes). To trace a proposed system tree, begin at any destination ("B" node) in card type 19 and use back nodes until the one reached is blank. Switch to card type 18 and obtain the back node for this "B" node. Next, go to card type 17 and continue to the home node ("A" node). The path is traced backwards. 4. Card Type 02 - System Link Times, Input to the Assignment 2 Program Card Columns Content 1- 2 Card type, "02" 3- 5 City code 6- 8 "A" node 12-14 "B" node (1) 19-20 Link time, X.X minutes 30-32 "B" node (2) 37-38 Link time, X.X minutes U8-50 "B" node (3) 55-56 Link time, X.X minutes 66-68 "B" node (4) 73-7^ Link time, X.X minutes 5o Card Type 24 - Skim Trees for Present or combination of Present and Proposed System Card Columns Content 1- 2 Card type, "24" 3- 5 City code 6- 8 Blank 9-11 Origin zone 12-14 Low destination zone 15 Blank 16-20 Driving time 1 (Low destination zone) 21-25 " 2 " .. +1 26-30 " 3 " " +2 XI- 13 Card Columns Content 31-35 Driving time 4 (Low destination zone +3 36-40 " 5 „ » +4 kl-k-5 « 6 " » +5 46-50 " 7 „ " +6 51-55 » 8 H » +7 56-60 •• 9 „ » +8 61-65 •i 10 n » +9 66-70 "11 .. » +10 71-75 " 12 » " +11 76-79 Identification (if any) 80 System type: ' 1" for present si fstem "2" for present and proposed system L. PROGRAM DESCRIPTION Prepared in January 1965 by Clyde E. Sweet, Jr. , Urban Development Branch, Urban Planning Division, Office of Planning, Bureau of Public Roads. XI-14 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XII - TRIP TABLE BUILDER PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 6 51 - IBM 1620 TRIP TABLE BUILDER PROGRAM Written by: Mr. Edwin D. Peterson, Georgia State Highway Department B. PURPOSE This program builds trip tables from variable format data card types 02, 03, 0^, or 05 and recodes the origin and destination subzones to zones (centroids) which are punched with their associated trip volumes. C. EQUIPMENT REQUIREMENTS This program is in SPS II for a *K)K IBM 1620 with card input-output, indirect addressing, branch not last card, and automatic divide special features,, D. PROGRAM RESTRICTIONS A maximum of 1,999 subzones can be recoded to a maximum of 375 zones, with 375 being the highest valid zone number. The original trip volumes can be rounded to five-position totals from either six-or seven-position totals, depending upon the setting of sense switch 1. The total volume that can be accumulated for any zonal movement is 99,999. Subzone equivalents (card type 9) should be sorted on the subzone number in ascending order — ' . There may be numbers missing in either the 1/ Ascending sequence is not mandatory for card type 9 in this program as it Is in the Trip End Summary program. five-digit subzone or the three-digit zone .(centroid) sequence, and the three-digit zones may be assigned at random. The three-digit zones (centroids) may be referenced by more than one five-digit subzone number; i.e., the same three-digit number may appear more than once. A trailer card is used with the type 9 cards. A centroid of zero (000) is not allowed. The range of three- digit centroids allowed is 001 through 200. E. OPTIONS AND SWITCH SETTINGS Option Switch Settings 1 - ON Volume totals are to be rounded from seven positions to five positions 1- OFF Volume totals are to be rounded from six positions to five positions. Use this switch setting unless there will be totals exceeding six digits. 2, 3, and k NOT USED PARITY - STOP 1/0 - STOP FLOW - PROGRAM F. OPERATING INSTRUCTIONS 1« Place all cards in the read hopper in the following sequence: a. Object program b. Card type 9 c. Card type 9 trailer card (99999 in subzone field) d. Card type 8, format card e. Survey card types 02, 03, 0^, or 05. Each group must be preceded with a card type 8 describing its format. 2o Clear storage. XII- 2 3o Press RESET. k. Press LOAD. 5. When machine halts, press start. 6. End of processing is identified by a typed message. G. MESSAGES AND PROGRAM HALTS Typed Message IMPOSSIBLE CARD TYPE NON-EXISTENT Machine Action Types out input area (Read + l) containing error and then halts . Types out input area (Read + l) containing error and then halts. EQUIVALENT CODE Types out message EXCEEDS 375 and then halts. NUMBER OF ORIGINAL SUBZONES EXCEEDS 1999 END OF SQUARE TRIP TABLE JOB Writes error message and halts. Explanation Card type read is not acceptable to this program. Run out cards, correct cards, reload all data and push console start and reader start. Card which was read contains a subzone code which is coded 99999 or is nonexistent. Run out card hopper, correct card in error, and reload corrected card and all cards following. Push console start and reader start to continue. If you can't correct the error, pull the job. Card which was read contains a zone equivalent which exceeds 375* R un ou t hopper, correct cards, and reload corrected card and all cards following. Push console start and reader start to continueo There is no restart from this error. Correct type 9 cards and reload program to begin again. End-of-Job. To run another job, reload program. xii-3 H. TIMING The time required in a test run to load the program, to read 739 type 9- cards and 226 survey cards and to punch trip tables for 8^- zones was 8 minutes. The time required for processing about 25,000 trip cards with 300 type 9 cards for 100 centroids was 58 minutes. I, METHOD The program begins by reading the subzone equivalents (type 9 cards). It then reads a type 8 format card and processes all the following survey data cards by converting the subzone number to the new zone (or centroid) number and inserting the number of trips in the zonal trip table. After reading all of the survey trip cards, the trip table (type 03 cards) is punched with the trips rounded according to the setting of program switch 1. If there is another card type 8 immediately following the previous survey trip cards, the program will process these cards into the same trip table. There can be any number of data cards sets (type 8 plus data cards), as long as the total numbers of trips do not exceed the five-digit maximum. The program also checks the card type and checks for consistency in city code. J. INPUT CARD TYPES 1. Card Type 8 - Format Card for Describing each survey trip card type Card columns should be blank if the information is not used in the card type being described. Card Columns Contents (c.c. = card column) 1 Card type, "8" 2- 4 City code, required 5- 7 Number of zones or centroids, required XII-4 Card Columns Contents (c .c. = card column) 8- 9 Survey card type described in this format card, only 02, 03, 04, or 05 accepted (required) 10-11 High order card column of origin subzone number, required 12-13 Low order c.c. of origin subzone number, required 14-15 High order c.c. of destination subzone number, required 16-17 Low order c.c. of destination subzone number, required 18-19 Card column containing trip purpose "from" (not required for trip table builder) 20-21 Card column containing trip purpose "to" (not required for trip table builder) 22-23 High order c.c. of trip expansion factor, required 24-25 Low order c.c. of trip expansion factor, required 26-27 High order c.c. of station operated. This should be coded so that the first number in this field is an actual digit of the station number. One-digit station numbers should be coded with a leading zero if there are also two-digit station numbers. 28-29 Low order c.c. of station operated 30-31 High order c.c. of station of entry or exit 32-33 Low order c.c. of station of entry or exit 34-35 Direction card column. Codes in survey cards will be : "1" for inbound "2" for outbound 36-37 Lowest external sector code. The sector code number has to be the left most digit in the subzone field of the survey cards A lowest external sector code of one digit is coded with a leading zero (required). xii- 5 Card Columns Content 38 Purpose number, required (1,2,3, or k) 39-80 Blank 2. Card Type 9 - Survey Subzone to Zone (centroid) equivalents Card Columns Content 1 Card type, "9" 2- 4 City code 5- 9 Subzone number, 5 digits 10-12 New centroid or zone number, 5 digits 13-80 Blank 3. Card Type 9 Trailer Card Card Columns Content 1 Card type, "9" 2- k City code 5- 9 99999 10-80 Blank K. OUTPUT CARD TYPE 1. Card Type 03 - Trip Volumes Card Columns Content 1- 2 Card type, "03" 3- 5 City code 6 Purpose 7- 8 Blank 9-11 Origin zone 12-14 Low destination zone XII-6 Card Columns Content 15 Blank 16-20 Volume 1 21-25 " 2 26-30 3 31-35 k 36-^0 5 kl-U-5 6 ^-50 7 51-55 8 56-60 9 61-65 » 10 66-70 N U 71-75 12 76-80 Identification (optional) L. PROGRAM DESCRIPTION This description was prepared in January 1965» by the Urban Development B ran ch, Urban Planning Division, Bureau of Public Roads XII-7 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XIII - ADD TRIP TABLES PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 652 - IBM 1620 ADD TRIP TABLES Written by: Mrs. Lamelle B. Hamner, Urban Planning Division, U.S. Bureau of Public Roads B. PURPOSE This program is written to add trip tables by purpose to form a total trip table. C. EQUIPMENT REQUIREMENTS This program is written in Fortran II for use on a 20K IBM 1620 with card input-output. D. PROGRAM RESTRICTIONS The maximum number of zones allowed is 400. The maximum number of trips from any one zone to any other zone is 99 > 999* Any number of purposes is allowed. A parameter card is necessary. Input data cards (card type 03 ) must be sorted on columns 9-H (zone of origin) for a low to high sequence. E. OPTIONS AND SWITCH SETTINGS 1. This program has two options which are controlled by card columns 11 and 12 of the parameter card. a, Option 1-A "1" punch in column 11 indicates that origin zone sequence must be unbroken. A "blank" in column 11 indicates that there' may be origin zones missing. XIII-1 b. Option 2 -A "1" punch in card column 12 indicates that the program will not punch zero trip tables for missing origin zones. A "blank" in this column will cause zero trip tables to be punched for missing zones. 2. Switch Settings: Program switches 1, 2, 3, and k NOT USED PARITY - STOP I/O - STOP FLOW - PROGRAM OPERATING INSTRUCTIONS 1. Clear storage. 2. Load cards in read hopper in the following order: a. Fortran object deck with subroutines b. Parameter card c. Type "03" cards (output of Trip Table Builder or Gravity Model Programs). Sorted on card columns 9-11. d. Trailer card with "99" in columns 1 & 2 3. Press LOAD. After the program is loaded the message "LOAD SUBROUTINES" is typed out. Press console START. ^. When the message "ENTER DATA" is typed out, press console START. 5o After the start of program execution, the parameter card is read and checked, also the number of zones to be processed. If either is incorrect, an error message will be typed out, see G -Messages and Program Halts. 6 Type "03" cards are checked for city code, origin zone, destination zone, and sequence. If any card is out of sort or is in error, an error message will be typed out. See G - Messages and Program Halts. XIII- 2 7. Output of the program will be another deck of "03" cards which will contain the total volume of trips from each zone to every other zone This deck can be differentiated from the input deck by a "1" in column 8. 8. After the- type "03" cards are punched the message "PROGRAM COMPLETE" will be typed out and the machine stops. MESSAGES AND PROGRAM HALTS Typed Message "A 2 IS NOT IN COLUMN 72." In- put card is typed. "PAUSE." Machine Action Types out the message, types out the param- eter card, and then comes to a pauseo Explanation The parameter card requires a 2 as identification in column 72. Correct the card, reload all data, and press start to continue. 'NUMBER OF ZONES Types out the message, IS TOO LARGE, types out the param- MAXIMUM IS 4-00." eter card, and then Input card is typed. comes to a pause. "PAUSE. " The program can accept only 4-00 zones as input. Correct the parameter card, reload all data, and press start to continue. If more than 4-00 zones, pull job. 'CARD IS IN ERROR." Input card is typed. "1 ERROR CARDS - JOB ENDED. STOP." Types out the message, types out the type 03 card causing the error, types out an additional message, and halts. Wrong card type - it is not a type 03 or a type 99 cardo Correct card, reload the program and begin again D "CARD OUT OF SORT. Types out the message, Card type 03 is out of sorto CARD IS IN ERROR. " Input card is typed. "1 ERROR CARDS - JOB ENDED. STOP." types out the card, types out additional message and halts. Correct the deck and reload the program to begin again D CARD IS IN ERROR. " Input card is typedo "1 ERROR CARDS - JOB ENDED. STOP." Types out the message, types out the card in error, types out additional message and halts o Low destination zone in type 03 card is in error. Correct the deck and reload the pro- gram to begin again. PROGRAM COMPLETE, CARDS READ, 00 CARDS NOT USED. " Types the message and halts o Successful run of the program. Reload program to run another jobo xin-3 NOTE: A blank , , in the above error messages signifies that the sequence number of the card which was read will be typed outo H. TIMING For a machine with the automatic floating point special feature, it took 10o2 minutes to load the program, add three purpose trip tables and punch the total trip table for 8^- zones. For a machine without the automatic floating point special feature, it took l4„4 minutes to load, add three purpose trip tables and punch the total trip table for 8A- zones. I. METHOD The program reads the parameter card and checks on the identification and the number of zones. The type 03 cards are read and checked for city code, for origin zone sequence and for low destination zone. When a type 03 card for the next origin zone is read, the summed trip table for the previous origin zone is punched. This procedure continues until the program complete message or an error message is typed. J. INPUT CARD TYPES Content Identification of card type to be processed, "03" required. City code Blank "1," if option 1 is selected, otherwise blank. "1," if option 2 is selected, otherwise blank. 1. Parameter Card Card Columns 1- 2 3- 5 6-10 11 XIII-4- Card Columns Content 13-27 Blank 28-30 Number of zones 31-59 Blank 60-62 "000" (required) 63-71 Blank 72 "2, e " parameter card identification 73-80 Blank 03 Cards - Output of Trip Table Builder Program Card Columns Content 1- 2 Card type, "03" 3- 5 City code 6 Purpose code 7- 8 Blank 9-U Origin zone 12-14 Low destination zone 15 Blank 16-20 Volume 1 21-25 " 2 26-30 3 31-35 h 36-40 5 k\-h c j 6 ^-50 7 51-55 8 xin-5 Card Columns Content 56-60 Volume 9 61-65 " 10 66-70 11 71-75 " 12 76-8O Identifi 3. Type 99 Card - Trailer Card Card Columns Content 1- 2 "99" 3-80 Blank K. OUTPUT CARD TYPES Type 03 Cards - Input to Card Columns Traffic Assignment Program 2. Content 1- 2 Card type, "03" 3- 5 City code 6- 7 Blank 8 "1" (identification for summed trip tables) 9-11 Origin zone 12-14 Low destination zone 15 Blank 16-20 Volume 1 21-25 " 2 26-30 3 31-35 4 36-40 5 41-45 6 XIII-6 Card Columns Content A-6-50 Volume 7 51-55 56-60 61-65 66-70 71-75 76-8O L. PROGRAM DESCRIPTION This description was prepared in January 1965 by Mrs. Lamelle B. Hamner, Urban Development Branch, Urban Planning Division, U.S. Bureau of Public Roads. 9 10 12 Identification (optional) XIII-7 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XIV - ASSIGNMENT 2 PROGRAM September 1965 A. IDENTIFICATION Deck No.: TA 653 - IBM 1620 ASSIGNMENT 2 PROGRAM - All-or- nothing or diversion assignment. Written by: Mr. William E. Roper, Traffic and Planning Division Mississippi State Highway Department B. PURPOSE This program accepts the link time deck from the Assignment 1 program and packed volume cards from the Add Trip Tables, Gravity Model, or Trip Table Builder programs. The Assignment 2 program builds the minimum path trees and accumulates the volumes on the links on the minimum time paths. The program can provide directional and nondirectional volumes on all links and turning volumes for up to 140 freeway node intersections. For a diversion assignment, a summary of the diversion data is also provided. C. EQUIPMENT REQUIREMENTS This program is written in SPS I for use on a 60K IBM 1620 with card input-output, indirect addressing, branch not last card, branch last card and automatic divide special features. D. PROGRAM RESTRICTIONS This program will accommodate a network with a maximum of 699 nodes, of which there is no maximum for zone centroids except that the number available is reduced by the number used for the proposed system. The pro- posed system is limited to a maximum of 140 nodes. The maximum time in a XIV- 1 minimum time path is limited to 99*^ minutes and the maximum time on any link is 9«9 minutes. A maximum of 100 links is allowed in a minimum time path and a maximum volume of 199 > 998 is allowed on each link. The maximum volume restriction is not checked by the program. E. OPTI O NS AND SWITCH 'SETTINGS All options are controlled by switch settings except for the range of centroids and nodes which is specified in card type 01 (parameter card). Switch Setting 2 - OFF 3 - ON 3 - OFF 4 - ON PARITY - STOP I/O - STOP FLOW - STOP Option Build trees and accumulate volumes for existing routes only. Build trees and accumulate volumes for existing and proposed routes. Use zones (centroids) and stations as intermediate nodes. Do not use zones (centroids). and stations as intermediate nodes. All-or-nothing assignment only. Card types 04, 05, 06, and 07 can be punched. Diversion assignment data are punched. The diversion assignment provides card types 04, 05, 06, 07, and 08. Input volumes are nondirectional such as the gravity model output. Assignment 2 output will also be nondirectional. Card types 04 and 06 will not be punched. Input volumes are directional such as trip table builder output. Assignment 2 output will also be directional. Card types 04, 05, 06, and 07 can be punched. F. OPERATING INSTRUCTIONS 1. Press Reset and Insert, then clear storage with check switches to PROGRAM by typing in - 16000 1000000 RS. 2. Press Instant Stop and then Reset after cycle has been completed. 3. Place the cards in the read hopper in the following order: a. Assignment 2 condensed program deck b. Card type 01 Co Card type 02 d. Card type 03 !±, Press LOAD 5. When machine halts after loading program, press START. 6. When the last card is in the read hopper, press READER START. 7. The end of the computer run is indicated by the type out of "PROCESSING COMPLETE." 8. The output cards, types 0^, 05, 06, and 07 (also type 08 for a diversion assignment) are run out of the hopper. Sorting the cards on card column 2 will place the cards in order by type. ERROR MESSAGES AND PROGRAM HALTS Typed Message Machine Action Explanation PROCESSING COMPLETE Halts End-of-Job. To process another system, place a new deck of cards containing card types 01, 02, and 03 in the read hopper o Press START and READER START. Refer to section F - Operating Instructions, Step 6. xrv-3 NOTE: ERROR MESSAGES AND RESTART PROCEDURES Errors are identified through programed error checks. The message "ER" and "the number representing the error" is typed out. The input area for the last card read (the card causing the error message) is also typed for each error condition. A correction can be typed into INPUT-79} if feasible, and a branch 0089^- will continue the program. If the error cannot be corrected by typing into the input area, run out the cards in the hopper, correct the cards, load all the data cards back into the read hopper, and branch to the start of the program. Typed Message ER 1 Machine Action Halts after typing input area Explanation Incorrect card type* only type 01, 02, or 03 accepted. Correct if feasible. See "NOTE" above. Halts after typing input area City code in input card is different than that on card type 01. Correct the input area, if feasible. See "NOTE" above. The city code is checked for consistencey by all programs. Halts after typing input area The time for the minimum path being built has exceeded the 99*^ minute maximum. See "NOTE" above. Halts after typing input area Table C is full, this table contains the nodes yet to be extended. This is an indi- cation that an error in the system of nodes is allowing looping. See "NOTE" above. Halts after typing input area The existing system minimum path tree contains more than 100 links. This could be a fault in the tree. xrv-^ Typed Message Machine Action Explanation ER 12 Halts after typing The proposed system minimum input area path tree contains more than 100 links. This could be a fault in the tree. ER 15 Halts after typing The origin and/or destination input area zone in the type 03 volume card is not within the range of nodes specified in card type 1. See "NOTE" above. H. TIMING A 366 node system with 84 zones required 45 minutes to build each tree load the trips and punch the output for an all-or-nothing (directional) assignment. I. METHOD The program first reads the type 01 and type 02 cards and edits for card type. If an all-or-nothing assignment is indicated, a tree is built for the origin zone specified in the first volume card, and the trips for this origin zone are accumulated. When the change in origin zone occurs, .a new tree is built. If the program switches call for a diversion assignment, both types of trees are built and trips are assigned to the freeway minimum time path trees based on the diversion curve percentage. These procedures are repeated until the last trips are assigned. The output for a directional assignment is then punched as type 04, 05, 06, and 07, cards „ Type 04 and 06 cards are not punched for a nondirectional assignment. Card types 06 and 07 are punched for freeway node numbers and card type 08 is punched only for a diversion assignment. yrv-5 J. INPUT CARD TYPES - Card columns not specified may be zero or blank 1. Card Type 01 - Parameter Card Card Columns Content 1-2 Card type, "01" 3- 5 City code 15-17 Lowest number in range of zone centroid nodes. Should be 001. 18-20 Highest number in range of zone centroid nodes. 33-35 Lowest number in range of station nodes 36-38 Highest number in range of station nodes 51-53 Lowest number in range of arterial nodes 54-56 Highest number in range of arterial nodes 69-71 Lowest number in range of freeway nodes (when used) 72-74 Highest number in range of freeway nodes (when used) 2. Card Type 02 - System Link Times Card Columns Content 1- 2 Card type, "02" 3- 5 City code 6- 8 "A" node 12-14 "B" node (l) 19-20 Link time, X.X minutes 30-32 "B" node (2) 37-38 Link time, X.X minutes 48-50 "B" node (3) 55-56 Link time, X.X minutes 66-68 "B" node (4) 73-74 Link time, X.X minutes xrv-6 3. Card Type 03 - Volume Cards Content Card Columns 1- 2 3- 5 6 Card type, "03 " City Code "Purpose" if card type 03 is output of Trip Table Builder. "1" if card type 03 is output of Add Trip Tables program. 9-11 Origin zone 12-14 Low destination zone 15 Blank 16-20 Volume 1 (Low destination zone) 21-25 • 2 • • • +1 26-30 • 3 • • 1 +2 31-35 • 4 • • ' +3 36-^0 • 5 • • ' +k kl-k5 • 6 • • ' +5 46-50 » 7 • • • +6 51-55 • 8 • • ' +7 56-60 ' 9 • • +8 61-65 • 10 • • ' +9 66-70 ' 11 • • ' +10 71-75 • 12 • • • +11 76-80 Identification (optional) XIV- 7 K. OUTPUT CARD TYPES - Card columns not specified may be zero or blank 1. Card Type 04 - Directional Link Volumes Card Columns Content 1- 2 Card type, "04" 3- 5 City code 6- 8 "A" node 12-14 "B" node (l) 15-20 Directional link volume 30-32 "B" node (2) 33-38 Directional link volume ^8-50 "B" node (3) 51-56 Directional link volume 66-68 "B" node (4) 69-74 Directional link volume Card Type 05 - Nondirectional Link Volumes Card Columns Content 1- 2 Card type, "05" 3-80 Same as card type 04 except that the volumes are nondirectional 3. Card Type 06 - Directional Turn Volumes These cards are obtained for only those nodes in the freeway node range. Card Columns Content 1-2 Card type, "06" 3- 5 City code node (location of interchange or intersection) XIV-8 Card Columns Content 9-11 "A" node (1) 12-14 "C" node (1) 15-20 Directional turn volume 27-29 "A" node (2) 30-32 "C" node (2) 33-38 Directional turn volume 45-47 "A" node (3) 48-50 "C" node (3) 51— 56 Directional turn volume 63-65 "A" node (4) 66-68 "C" node (4) 69-74 Directional turn volume A directional turning movement is identified by a combinationa of "A, B, and C" nodes. The "B" node locates the interchange. The "A" node to the "C" node indicates the direction of the turning movement. 4. Card Type 07 - Nondirectional Turn volumes These cards are obtained for only those nodes in the freeway node range. Card Columns Content 1- 2 3-80 Card type, "07" Same as card type 06 except that the volumes are nondirectional. XIV-9 5. Card Type Diversion Assignment Data These cards are punched for a diversion assignment only, to check the assignment for reasonableness. Card Columns Content 1-2 Card type, "08" 3- 5 City code 9-11 Origin zone 12-14 Destination zone 15-20 Present volume 27-29 . Time to traverse present path 30-32 Time using freeway path 33-38 Present remaining volume 39-4^- Present diverted volume 45-47 Time on present path without common link times 48-50 Time using freeway path without common link times 63-65 Back node of intermediate link used for the freeway path 66-68 H B" node of destination of the inter- mediate link used for the freeway path 72-76 Time ratio of freeway route time to present route time 78-80 Percentage of traffic diverted L. PROGRAM DESCRIPTION Prepared in January I965, by Clyde E. Sweet, Jr. , Urban Development Branch, Urban Planning Division, Office of Planning, Bureau of Public Roads <> U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XV - UPDATE SKIM TREES PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 65^ - IBM 1620 UPDATE SKIM TREES PROGRAM Written by: Mr. Edwin D. Peterson, Georgia State Highway Department and revised by Mr. Clyde E. Sweet, Jr., Urban Planning Division, U.S. Bureau of Public Roads B. PURPOSE This program prepares the updated skim trees. The program combines the intrazonal and interzonal driving times with the terminal times and develops the total traveltime from each zone to all other zones in the study area. C. EQUIPMENT REQUIREMENTS This program is written in Fortran II for use on a 20K IBM 1620 with card input- output. D. PROGRAM RESTRICTIONS This program will process a maximum of 200 zones. E. OPTIONS AND SWITCH SETTINGS 1. There are no options 2. Switch Settings: PARITY - STOP I/O - STOP FLOW - PROGRAM OPERATING INSTRUCTIONS 1. The type 31 cards must be in sort by origin zone. The type 2k cards must be in order as punched from the Assignment 1 program - major sort on origin zone and minor sort on low destination zone. XV- 1 2. Clear storage with switches set to program. 3. Press RESET. 4. Place cards in the read hopper in the following order: a. Fortran object deck with subroutines b. Parameter card c. Type 31 cards d. Type Zh cards 5. Press LOAD to begin. 6. The message "LOAD SUBROUTINES" will be typed out. Press START. 7. The message "ENTER DATA" will be typed out. Press START to begin processing. 8. Press READER START for the last type Zh card. 9. Error halts may occur during processing. 10. After all cards are processed and the output is punched, the job complete message will be typed out and the computer will halt. G. MESSAGES AND PROGRAM HALTS Typed Message "NUMBER OF ZONES IN ERROR, CORRECT PARAMETER CARD, RELOAD DATA, PRESS START TO BEGIN AGAIN. PAUSE." "NOT CARD TYPE 31, CORRECT CARD, RELOAD ALL DATA, PRESS START TO BEGIN AGAIN. PAUSE." "NOT CARD TYPE Zk, CORRECT CARD, RELOAD halts TYPE Zk CARDS, CLEAR PUNCH HOPPER, PRESS START TO PROCEED. PAUSE." Machine Action Types message and halts. Types message and halts. Types message and Explanation Check the number of zones used in the parameter card. The maximum permissible is 200. Press start to begin by reading the parameter card again. Check card order, correct card, reload parameter card and all data, press start to begin again. Correct card, reload type 2k cards, run out and restart the punch, press console start to continue. Typed Message Machine Action Explanation "JOB COMPLETE. Types message and Successful run. Reload STOP." halts. program to make another run. H. TIMING To load the' program, read the type 31 and 2k data cards, and punch the updated skim trees for 84 zones required 21 minutes. I. METHOD The terminal time and intra zonal time for each zone are read in from the deck of type 31 cards which have previously been sorted on the origin zone. This information is stored. The minimum path driving time from each zone to all other zones is read in from the deck of type 2k- cards. Halts will occur on certain errors. See G - Messages and Program Halts. The appropriate terminal times are added to the intrazonal driving time and interzonal driving time to arrive at the total minimum path travel- time for all zone- to- zone movements in the study area. J. INPUT CARD TYPES 1. Parameter Card Card Columns Content 1-3 Total number of zones (maximum = 200) k— 6 City code 7- 9 Number of type 2k- cards per zone of origin times 12 10-80 Blank Type 31 cards - Terminal and Intrazonal time Card Columns Content 1-2 Card type, "31" 3- 5 City code XV- 3 Card Columns 6- 8 9-13 14-18 19-80 Content Origin zone Terminal time xxxx.x Intrazonal time xxxx.x Blank Type 24 cards - Skim Trees for Present or Proposed System Card Columns 1- 2 3- 5 6- 8 9-11 12-14 15 16-20 21-25 26-30 31-35 36-40 41-45 46-50 51-55 56-60 61-65 66-70 71-75 76-79 Content Card type, "24" City code Blank Origin zone Low destination zone Blank Driving time 1 2 3 4 5 Identification, if any System type "1" = present "2" = present and propo sed XV-4 K. OUTPUT CARD TYPE Type 2k cards - Updated Skim Trees for Present or Proposed System Card Columns Content 1- 2 Card type, •24" 3- 5 City code 6- 7 Blank 8 "1" - Identification for updated skim trees 9-11 Origin zone 12-14 Low destination zone 15 Blank 16-20 Traveltime 1 21-25 „ 2 26-30 » 3 31-35 M 4 36-40 ,, 5 41-45 „ 6 U6-50 » 7 51-55 » 8 56-60 ■ 9 61-65 ,. 10 66-70 „ 11 71-75 » 12 76-79 Identificat Lon, if any 80 System type XV- 5 "1" = Present "2" = Present and Proposed L. PROGRAM DESCRIPTION This description was prepared in January 19&5 > "by "the Urban Development Branch, Urban Planning Division, U.S. Bureau of Public Roads. U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XVI - TRIP LENGTH FREQUENCY PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 655 - IBM 1620 TRIP LENGTH FREQUENCY PROGRAM Written by: Mr. Edwin D. Peterson, Georgia State Highway Department B. PURPOSE This program computes and punches the following for each one-minute increment of traveltime: The trip volume, the percentage that the trip volume is of the total volume, and the minutes of travel. C. EQUIPMENT REQUIREMENTS This program is written in SPS II for use on a 20K IBM 1620 with card input-output, indirect addressing, branch last card, and automatic divide special features. D. PROGRAM RESTRICTIONS This program can accept trip volumes for a maximum of ^4-00 centroids (zones). Trip volumes can be accumulated for each one-minute increment from 1 to 99 minutes of traveltime. The maximum accumulated volume is 999,99' E. OPTIONS AND SWITCH SETTINGS 1. There are no options; program switches are not usedo 2. Switches: PARITY - STOP I/O - STOP FLOW - PROGRAM XVI- 1 F. OPERATING INSTRUCTIONS 1. Place updated skim trees, type 2k cards, in front of type 03 cards and sort on origin zone (c.c. 9-H)« 2. Place the cards in the read hopper in the following sequence: a. Condensed program deck b. Cards sorted as in step 1. 3. Clear storage. k. Press RESET. 5o Press LOAD. 6. When machine halts press CONSOLE START. 7. When last card is in read hopper press READER START. 8. End of processing is identified by an "END OF JOB" type out. G. MESSAGES AND PROGRAM HALTS Typed Message "IMPOSSIBLE INPUT CARD TYPE." "VOLUME CARDS MISSING FOR CENTROID" Origin zone (centroid no.) is typed out. Machine Action Types message and halts. Types message and centroid number, then halts. Expl a nation The card read was not a type 2k or a type 03 . Correct cards, reload all data, press start to begin again. Volume cards are missing for this zone. If this is all right, press start. If this condition is incorrect, correct the input data, reload the program and begin again. "DRIVING TIME EXCEEDS 99 MINUTES." Types an 16 digits, and halts additional 16 digits. Types message, types There is a time in a type 2k card which exceeds 99 minutes. The digits typed are the last digit of the city code, three zeros, the origin zone, the low desti- nation zone, a zero, and the time. This is a nonrecoverable halt. Check out the skim trees and run the job over. "END OF JOB" Types message, and halts. End- of- job. -To run another job, reload the program. H. TIMING The time required for loading the program, reading all data cards for 84 zones, and punching the output for 22 minutes of traveltime was 6 minutes. I. METHOD The program reads the updated skim trees for one origin zone and sets up a table of time increments for the skim tree times rounded to the nearest whole minute. The trips for this origin zone (centroid) are then read and summed under the appropriate time increments. These procedures are repeated for all origin zones. The percentage of total trips and the minutes of travel are calculated for each time increment. After the final calculation, the type 37 card is punched for each one-minute time increment. J. INPUT CARD TYPES 1 Card type 24 - Updated Skim Trees for Present or Present and Proposed System Card Columns Content 1- 2 Card type, "24" 3- 5 City code 6- 7 Blank 8 "1" Identification punch 9-U Origin zone 12-14 Low destination zone 15 Blank 16-20 Driving time 1 21-25 " " 2 26-30 " " 3 31-35 " " 4 36-40 •• " 5 XVI-3 Card Columns Content 41-45 Driving time 6 46-50 " 7 51-55 » 8 56-60 " 9 61-65 M 1Q 66-70 " 11 71-75 " 12 76-79 Identification , if any 80 System type "1" "2" = Present = Present and proposed I type 03 - - Trip Volumes Card Columns Content 1- 2 Card type, "03" 3- 5 City code 6 Purpose 7- 8 Blank 9-11 Origin zone 12-14 Low destination zone 15 Blank 16-20 Volume 1 21-25 " 2 26-30 3 31-35 4 36-40 5 41-45 6 XVI-4 Card Columns Content ^6-50 Volume 7 51-55 " 8 56-60 9 61-65 " 10 66-70 11 71-75 12 76-8O Identification (optional) K. OUTPUT CARD TYPE Card type 37 - Trip Length Frequency Card Columns Content 1- 2 Card type, "37" 3- 5 City code 6 Blank 7- 8 Time increment in minutes 9-H1, Number of trips 15-19 Percent of total, xxx.xx 20-29 Minutes of travel 30-80 Blank L. PROGRAM DESCRIPTION This description was prepared in February 1965» by the Urban Development Branch, Urban Planning Division, Bureau of Public Roads xvi- 5 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XVII - TRIP END SUMMARY PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 656 - IBM 1620 TRIP END SUMMARY PROGRAM Written by: Mr. Don Winter, Georgia State Highway Department B. PURPOSE This program recodes five-digit origin and destination subzone codes into corresponding three-digit zone or centroid codes and computes the interzonal trips produced and attracted, intrazonal trips, total productions, total attractions and total trip ends for each production zone or centroid for all three trip purposes as well as for total trips, C. EQUIPMENT REQUIREMENTS This program is written in SPS II for use on a ^OK IBM 1620 with card input-output, indirect addressing, and automatic divide special features. D. PROGRAM RESTRICTIONS This program is not relocatable. It will process cards containing a maximum of 1,000 five-digit subzone codes and recode them into a maximum of 200 zones or centroids. The total accumulated trip volume for any zone or centroid is 99,999. Subzone equivalents (card type 9) must be sorted on the subzone number in ascending order. There may be numbers missing in either the five-digit subzone or three-digit zone (centroid) sequence, and the three-digit zones may be assigned at random. Three-digit zones may be referenced by more XVII- 1 than one five-digit subzone number; i.e Q , the same three-digit number may- appear more than once. No trailer card is used. A zone (centroid) of zero (000) is not allowed. The range of three-digit centroids allowed is 001 through 200. E. OPTIONS AND SWITCH SETTINGS There are no options in this program. Switch setting 1, 2, 3? and h not used PARITY - STOP 1/0 - STOP FLOW - PROGRAM F. OPERATING INSTRUCTIONS 1. Press RESET and INSERT, then clear storage by typing in 160001000000RS. 2. Press INSTANT STOP and RESET. 3. Place cards in read hopper in the following order: a. Condensed program deck b. Type 9 cards in ascending order. ( DO NOT USE the 99999 trailer card used with the Trip Table Builder). c. Type 8 card d. Card type 02, 03, 04, or 05. Card groups may be entered in any order but only one type will be processed at a time. The type 8 format card must precede each type of survey card. e. A card with a record mark in card column 1 (0, 2, and 8 punches). 4. Push LOAD; when machine halts, push console START. XVII-2 5. At times the program may appear to be in an endless loop. Without errors the program will process until end-of-job message is typed out, (" CONS UMAT ION. ") G. MESSAGES AND PROGRAM HALTS Typed Messages Machine Action Explanation ERR 1 Unrecoverable halt Invalid card type. Correct cards, reload program and begin again. ERR 2 Unrecoverable halt Subzone not shown in card type 9» This is either an error in the survey card or there is a type 9 card missing. Correct error and begin again. ERR 3 Unrecoverable halt Card type 9 out of sequence. Resequence type 9 cards and begin again. H. TIMING The time required in a test run to load the program, to read 739 type 9 cards and 226 survey cards and to punch the trip end summary cards for 84 zones was 5 minutes. I. METHOD The program starts by reading the subzone equivalents (type 9 cards). A type 8 format parameter card is then read and each of the survey cards in this format is read and processed by converting the subzone number to the new zone (centroid) number and determining the zone of production and attraction for each survey card. The purpose of the survey trip card is determined from the purpose "to" and purpose "from." A purpose "to" or "from" of zero places the trips in the home based trip purpose category. Similarly, if both purpose "to" and "from" are nonzero, the trips are XVII-3 placed in the nonhome based trip purpose category. If the trips are home based, the purpose "to" and "from" are checked for a 1. The trip cards with a 1 are placed in the home based work trip category; all other home based trips would be placed in the other home based (nonwork) trip category. At this point, the number of trips from each survey card is placed in the appropriate production and attraction zone tables. After all the type 8 cards and the following survey cards have been processed, the type 30 cards are punched for each purpose. The purposes are punched in sequence by production zone number in the following order: 1, 2, 3 » and 4 (c.c. 39-^-0). Purpose 1 is work, purpose 2 is nonwork, purpose 3 is nonhome based, and purpose U- is the total. Centroids with zero volumes will have a type 30 card punched; centroids not referenced in type 9 cards (e.g., numbers not used as centroids but falling within the sequence of centroid numbers) will have a card type 30 punched. A card type 30 must be present for each centroid, starting with 001 and going to the highest number, for use with the Gravity Model program. J. INPUT CARD TYPES 1. Card type 8 - Format Card for describing each Survey Trip Card type. Card columns should be blank if the information is not used in the card type being described. Card Columns Content (c.c. = card column) 1 Card type, "8" 2- k City code, required. 5- 7 Number of zones or centroids, required. 8- 9 Survey card type described in this format card. Only 02, 03, 04, or 05, accepted. XVII-4 Card Columns 10- -11 12- ■13 Ik- ■15 16- •17 18- •19 High order card column of origin subzone number, required. Low order c.c. of origin subzone number, required. High order c.c. of destination subzone number, required. Low order c.c. of destination subzone number, required. Card column containing trip purpose "from" (required). Card column containing trip purpose "to" (required). Note: The purposes "to" and "from" used in this program are as follows: Code Purpose 1 2 Work Business 3 Medical-dental 4 School 5 Social-recreational 6 Change travel mode 7 Eat-meal 8 Shop 9 Serve passenger Home High order c.c. of trip expansion factor required. 22-23 24-25 Low order c.c. of trip expansion factor, required. 26-27 High order c.c. of station operated. This should be coded so that the first number in this field is an actual digit of the station number. One-digit station numbers should be coded with a leading zero if there are also two-digit station numbers. 28-29 Low order c.c. of station operated. 30-31 High order c.c. of station of entry or exit. XVII- 5 Card Columns Content 32-33 Low order c.c. of station of entry or exit. 3^-35 Direction card column. Codes in survey cards will be: "1" for inbound "2" for outbound 36-37 Lowest external sector code. The sector code number has to be the left most digit in the subzone field of the survey cards. A lowest external sector of one digit is coded with a leading zero, required. 38 , Purpose number, not required. 39-80 Blank 2. Card type 9 - Subzone to Zone (centroid) equivalent cards Card Column s Content 1 Card type, ,, 9 ,, 2- k City code 5- 9 Subzone number, 5 digits 10-12 New centroid or zone number, 3 digits 13-80 Blank K. OUTPUT CARD TYPE 1. Card type 30 (One deck punched for each purpose) Card Columns Content 1- 2 Card type, "30" 3- 5 City code 6- 8 Zone of production 9-13 Interzonal trips produced Card Columns Content 1*1-18 Interzonal trips attracted 19-23 Intra zonal trips 24-28 Total productions 29-33 Total attractions 34-38 Total trip ends 3.9-^0 Trip purpose - 01, 02, 03, (04 is total of purpose 01 03) 41-80 Blank PROGRAM DESCRIPTION This description was prepared in February 1965? by the Urban Development Branch, Urban Planning Division, Bureau of Public Roads. XVII- 7 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XVIII - GRAVITY MODEL PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 657 - IBM 1620 GRAVITY MODEL PROGRAM Written by: Mr. Don Winter, Georgia State Highway Department B. PURPOSE This program distributes trips according to the gravity model formula for one purpose at a time. The program computes the trip volumes, zonal attractions and a trip length frequency distribution. C. EQUIPMENT REQUIREMENTS This program is written in SPS II for use on a 20K IBM 1620 with card input-output, indirect addressing and automatic divide special features. D. PROGRAM RESTRICTIONS The program can accommodate a maximum of 200 zones and 99 one-minute time increments. Input type 2h cards must be sorted as follows: Major sort on origin zone and minor sort on low destination zone, in ascending order » This is the same order in which they are punched from the Assignment 1 program. E. OPTIONS AND SWITCH SETTINGS The punching of the gravity model trip tables is controlled by switch 1. Switch Settings Option 1 - OFF Do not punch gravity model type 03 trip tables 1 - ON Punch gravity model type 03 trip tables PARITY - STOP I/O - STOP O FLOW - STOP F. OPERATING INSTRUCTIONS 1. Clear storage with check switches set to PROGRAM. 2. Place cards in read hopper in the following order: a. Condensed program deck b. Type 08 card (gravity model parameter card) c. Type 30 cards (productions and attractions, one purpose only) d. Type 31 cards (terminal and intrazonal times) e. Type 32 cards (traveltime factors for the same purpose as type 30 cards) f. Type 33 cards (if K factors are used, they must be inserted in the type 2h deck immediately before the origin zone to which they apply). g. Type 2k cards (must be in sort - major sort origin, minor sort low destination) h. Blank card with record mark (0, 2, and 8 punches) in column 1 3. Press LOAD to begin loading program. k. After loading the machine halts. 5. Check switches and hit console START. 6. An end-of-job message will be typed at the end of a successful run. For explanation of typed messages, see G- Messages and Program Halts. G. MESSAGES AND PROGRAM HALTS Typed Message Machine Acti o n Explanation "ER 1" Types message and Invalid card type. There is no halts. recovery from this halt. Correct cards and reload the program to begin again. "CONSUMATION" Types message and End-of-Job. To run another halts. purpose, change data cards and reload the program. H. TIMING The time required to run the gravity model for 22 time increments and to punch the trip tables for 84 zones is 25 minutes. There were no K factors used in this operation. I. METHOD The program initially reads the parameter card (type 08) and sets up the counters and test areas. Next the production and attraction cards (one type 30 for each centroid) for one purpose are read. Before processing finally begins, the intrazonal and terminal time cards are read (type 31) and the traveltime factor cards are read (type 32). The application of the gravity model formula begins by reading the K factor cards (type 33? if any) and then the skim tree cards (type 24) for the first zone. The computations begin with the calculation of the total traveltime. The appropriate travel- time factors, productions, and attractions are then used to calculate the trips from the first production zone to all attraction zones. K factors are also applied at this time to any specified attraction zones. On completion of the calculations for the first production zone, K factors and skim trees are read for the next production zone. The trips for each succeeding production zone are computed in the same manner until all centroids have been processed. After the trips have been calculated for each production zone, the trip tables are punched if the option has been specified. The next output are the model attractions (card type 30a). The attractions are then adjusted and punched into card type 30b. The trip length frequency distribution is punched into type 35 cards. An end-of-job message is typed after the last output card has been XVIII-3 punched. Before making the next run of the gravity model, adjust the traveltime factors and punch the adjusted traveltime factors into a new set of type 32 cards. All data except the type 32 cards would remain the same for each calibration of the gravity model. The type 30b cards may be used as input to a succeeding run of the gravity model program in order to balance the model attractions for the same set of traveltime factors. J . INPUT CARD TYPES 1. Card Type 08 ■ - Gravity Model Parameter Card Card Columns Content 1- 2 Card type, "08" 3- 5 City code 6- 8 Number of centroids 9-80 Blank 2. Card Type 30 - Productions and Attractions, one purpose for each run Card Columns Content 1- 2 Card type, "30" 3- 5 City code 6- 8 Zone of production 9-13 Interzonal trips produced 14-18 Interzonal trips attracted 19-23 Intra zonal trips 24-28 Total productions 29-33 Total attractions 34-38 Total trip ends 39-^0 Trip purpose 41-80 Blank XVIII-4 Card type 31 - Terminal and Intrazonal times Card Columns Content 1- 2 Card type, "31" 3- 5 City code 6- 8 Origin zone 9-13 Terminal time XXXX.X 14-18 Intrazonal time XXXX.X 19-80 Blank 4. Card type 32 ■ - Traveltime Factors for the same purpose as card type 30 Card Columns Content 1- 2 Card type, "32" 3- 5 City code 6- 8 Time in minutes 9-13 Traveltime factors XX. XXX 14-80 Blank Card type 24 - Skim Trees for present or present and proposed system Card Columns Content 1- 2 Card type, "24" 3- 5 City code 6- 8 Blank 9-11 Origin zone 12-14 low destination zone 15 Blank 16-20 Driving time 1 21-25 11 ii 2 26-30 ,. 3 xviii- 5 Card Columns Content 31-35 Driving time h 36-^0 ,. 5 if-1-^5 " 6 ^-50 " 7 51-55 ,, 8 56-60 .. 9 61-65 •1 10 66-70 .. n 71-75 , 11 1. 12 76-79 Identification 80 System type "1" = Present "2" = Present and proposed type 33 - - K Fa 3tors, optional Card Columns Content 1- 2 Card type, "3" 3- 5 City code 6-15 Blank 16-18 Attraction zone number 19-22 K factor XXX. X. 23-80 Blank NOTE: All K factors are considered 001.0 unless each group (group by origin) of type 2h cards is preceded by appropriate card type 33 • No origin code is necessary, and only those K- factors other than 001.0 must be represented. XVIII-6 K. OUTPUT CARD TYPES 1. Card type 03 - Trip Cards Card Columns Content 1- 2 Card type, "03" 3- 5 City code 6- 8 Blank 9-11 Origin zone 12-14 Low destination zone 15 Blank 16-20 Volume 1 21-25 2 26-30 3 31-35 4 36-40 5 41-45 6 46-50 7 51-55 8 56-60 9 61-65 10 66-70 11 71 75 12 76-80 Identification (optic xviii-7 Card type 30a - Gravity M del output only, productions and model attractions Card Columns Content 1- 2 Card type, "30" 3- 5 City code 6- 8 Zone of production 9-23 Blank 2^-28 Total productions 29-33 Total attractions, unadjusted 3^-80 Blank NOTE: The model (unadjusted) attractions are always punched into the first deck of the card type 30 gravity model output. Card Type 30b - Gravity Model Output only, productions and adjusted attractions Card Columns Content 1- 2 Card type, "30" 3- 5 City code 6- 8 Zone of production 9-23 Blank 2^-28 Total productions 29-33 Total attractions, adjusted 3^-80 Blank Type 35 ■ - Trip Length Frequency Distribution Card Columns Content 1- 2 Card type, "35" 3- 5 City code 6- 8 Time in minutes XVIII-8 Card Columns Content 9 Blank 10-14 Number of trips 15-19 Percent of total trips XX. XX 20-21 Blank 22-29 Minutes of travel 30-80 Blank L. PROGRAM DESCRIPTION This description was prepared in February 19&5* by Clyde E. Sweet, Jr. Urban Development Branch, Urban Planning Division, U.S. Bureau of Public Roads. U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XIX - TRIP COMPARISON PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 658 - IBM 1620 TRIP COMPARISON PROGRAM Written by: Mrs. Rose M. Hall, Urban Planning Division, U.S. Bureau of Public Roads B. PURPOSE This program compares the assigned gravity model trips with the assigned origin-destination survey trips. The program provides dif- ferences between the gravity model link volumes and 0-D survey link volumes. A statistical analysis of differences is provided for each volume group. C. EQUIPMENT REQUIREMENTS This program is written in Fortran II for use on a 60K IBM 1620 with card input-output, automatic divide, and automatic floating point special features. A different version of this program is available for users without the automatic floating point special feature (hardware). D. PROGRAM RESTRICTIONS A maximum of 700 nodes can be processed by this program. The nodes must be consecutively numbered beginning with one The number of volume and difference groups is limited to 15. These groups must range from low to high. E. OPTIONS AND SWITCH SETTINGS Options 1. Option to punch table 1. XIX-1 If table 1 is desired, it must be specified in the parameter card with a "3" coded in column 3» See J - Input Card Types. - 2o Option to punch only the portions of table 2 with frequencies. If -this option is not selected, all difference groups will be punched for each volume group, with zeros in the difference groups without entries. This option is specified with a "5" coded in column 5 of the parameter card. See J - Input Card Types. Switch Settings - program switches not used PARITY - STOP I/O - STOP FLOW - PROGRAM OPERATING INSTRUCTIONS 1. Clear storage. 2. Press RESET. 3. Load cards in the read hopper in the following order: a. Fortran object deck with subroutines b. Identification card (required) c. Parameter card d. Difference group card e. Volume group card f . Type 05 cards from 0-D assignment g. Type 99 trailer card h. Type 05 cards from G-M assignment i. Type 99 trailer card k. Press LOAD. XIX-2 5- When the message console START. 'LOAD SUBROUTINES" is typed out, press When the message START. 'ENTER DATA" is typed out, press console 7. Press punch START. 8. There will be a long hesitation before program execution begins. After the program begins execution, the first three cards are typed out: a. Parameter card b. Difference group card c. Volume group card If any of these cards are out of order or incorrectly coded, the appropriate error message is typed out and the machine will pausec See G - Error Messages and Program Halts 9. The program punches the output for the selected options and writes out the totals and an end-of-job message upon completion of the punching. G. ERROR MESSAGES AND PROGRAM HALTS Typed Message "ABOVE CARD IS IN ERROR, INCORRECT IDENTIFICATION OR SEQUENCE. CORRECT CARD, RELOAD DATA, PRESS START TO CONT." Machine Action Types out the card, then a message, and then it pauses. Explanation Correct card, reload all data, and press start to continue. "NOT A NO. 5 CARD. STOP." Unrecoverable halto Pull job. Check the assigned link volumes. All cards in the deck are not type 05 cards. Reload program to run again. "END OF PRO- GRAM. STOP." Halt Processing is complete. List all output cards to obtain the statistical data. Reload the program to make the next comparisons. XH-3 H. TIMING To load the program and run the comparisons for a 25 node test assignment required 8 minutes „ To load the program and run the comparisons for a 366 node assignment required 22 minutes. I. METHOD 1. The information for the origin-destination survey is read in from the first deck of type 05 cards and stored. This information includes for each A node the connecting B nodes and the corresponding assigned link volumes. The second deck of type 05 cards (gravity model data) is read in one card at a time. Each card is checked against the corresponding 0-D card for the A node number Then the first B node number is checked and the difference between the 0-D and G-M link volumes for this B node is computed. If option 1 has been specified, the node to node comparison is punched. This card -will include A node, B node, 0-D volume, G-M volume and the difference. The difference between the two link volumes is then stored in the proper difference group table for the designated volume group, determined by the 0-D link volume. At the same time additions are made to the following tables within the volume group: a e Frequencies for difference group b Sum of differences squared c. Total 0-D link volumes d. Total G-M link volumes e Total of frequencies for volume group XTX-4 This procedure is executed for each of the B nodes on the card. Then the next G-M type 05 card is read in and the process is repeated, until all G-M type 05 cards have been processed. 2. The second portion of the program consists of computing the statistics for the analysis of differences between the origin-destination survey and gravity model link volumes. The program develops and punches the frequency distribution of the differences by volume group. For each volume group it calculates the following: total frequency, total difference, sum of the squares of the differences, mean difference, root-mean- square error of the differences, percent root-mean-square error, total 0-D trips, and total G-M trips If option 2 has been specified, only the groups with frequencies will be punched, otherwise, all difference groups will be punched for each volume group, with zeros in the difference groups without entries. J. INPUT CARD TYPES 1. Identification Card Card Columns Content 1-80 Can contain anything desired, will be used as identification heading on table 1. Parameter Card Card Columns 1- 2 3 k 5 Content Blank "3" for option to punch table 1, otherwise blank. Blank "5" for option to punch only that portion of table 2 with entries, otherwise blank. xn-5 Card Columns Content 6-12 Blank 13-16 Number of nodes in network 17-18 Blank 19-20 Number of difference groups 21-22 Blank 23-2^ Number of volume groups 25-79 Blank 3. Difference Group Card - Comparison Program Only Card Columns Content 1- 5 First difference group 6-10 Second M « 11-15 Third " ' 16-20 Fourth " » 21-25 Fifth " » 26-30 Sixth " • 31-35 Seventh " ' 36-^0 Eighth " • M-^-5 Ninth » » ^-50 Tenth " • 51-55 Eleventh " • 56-60 Twelth " ' 61-65 Thirteenth " ■ 66-70 Fourteenth " ' xix-6 Card Columns Content 71-75 Fifteenth difference group 76-79 Blank 80 "3" Volume Group Card - Comparison Program Only Card Columns Content 1- 5 First volume group 6-10 Second " " 11-15 Third " " 16-20 Fourth " " 21-25 Fifth " " 26-30 Srxth " " 31-35 Seventh " " 36-^0 Eighth » *n-45 Ninth " ^6-50 Tenth " " 51-55 Eleventh " " 56-60 Twelfth " 61-65 Thirteenth " " 66-70 Fourteenth " " 71-75 Fifteenth " " 76-79 Blank 80 »l\,» XIX-7 5o Card Type 05 - Origin-Destination Nondirectional Link Volumes Card Columns Content 1- 2 Card type, "05" 3- 5 City code 6- 8 "A" node 9-H Blank 12-14 "B" node (l) 15-20 Nondirectional link volume 21-29 Blank 30-32 "B" node (2) 33-38 Nondirectional link volume 39-^7 Blank ^8-50 "B" node (3) 51-56 Nondirectional link volume 57-65 Blank 66-68 "B" node (b) 69-7^ Nondirectional link volume 75-80 Blank This deck must be followed by a trailer card with "99" in columns 1 & 2. 6. Card Type 05 - Gravity Model Nondirectional Link Volumes Card Columns Content (Same format as for Origin-Destination Nondirectional Link Volumes) This deck must be followed by a trailer card with "99" in columns 1 & 2. K. OUTPUT CARD TYPES The output cards from this program are not used as input to any other program. They must be listed in the same order as punched to obtain printed tables. See chapter VI, tables VI- 1 and VI-2, for examples of the printed output L. PROGRAM DESCRIPTION This description was prepared in January 19&5 by Mrs. Rose M. Hall, Urban Development Branch, Urban Planning Division, U.S. Bureau of Public Roads. XIX-9 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS CHAPTER XX - FRATAR TRIP DISTRIBUTION PROGRAM September 1965 A. IDENTIFICATION Deck No. : TA 659 - IBM 1620 FRATAR TRIP DISTRIBUTION PROGRAM Written by: Original written by Terry J. Puckett, Mississippi State Highway Department. Revised by Alabama State Highway Department, October 1964. Revised by William E. Roper, Mississippi State Highway Department, December 1964. Revised by Clyde Eo Sweet, Jr., Bureau of Public Roads, ' June 1965. B. PURPOSE This program utilizes the Fratar trip distribution procedure to obtain trip volumes for a future time period. C. EQUIPMENT REQUIREMENTS This program was written in SPS I and revised in SPS II for use on a 60K IBM 1620 with card input-output, automatic divide, and indirect address- ing special features. This program is not relocatable. A different version of the Fratar program is available as deck number TA659X. This program requires the additional special features (instructions) TNS and TNFo This program can process one additional zone<> This program is also not relocatable. Either program may be modified to process a lesser number of zones on a machine of smaller core size. The storage requirements for a- particular number of zones can be calculated by the following equation: Storage Locations - 10,680 +2.5 (No. of Zones) (No. of zones - 1). XX- 1 D. PROGRAM RESTRICTIONS A maximum of 140 zones will be processed by this program. The zones must be consecutively numbered and there may not be a zone number of zero. The maximum nondirectional volume for each zone pair is 99 > 999. Growth factors must be greater than zero and can not exceed 99*999. E. OPTIONS AND SWITCH SETTINGS There is only one option controlled by a switch setting. The remaining options are controlled by typing in from the typewriter (see Operating Instructions) or by changing the program. The machine size requirement may be changed by revising the card identified as page 03, line 040, see also Equipment Requirements. Switch Settings 1 - ON 1 - OFF Option Punch the type 03 card trip tables for each approximation. Punch the type 03 card trip tables for the last approximation only . Switch 2 must be on at all times. 2 - ON 3 and 4 - NOT USED PARITY - STOP I/O - STOP FLOW - STOP F. OPERATING INSTRUCTIONS 1. Press Reset and Insert, then clear storage with check switches to PROGRAM by typing in - 160001000000RS. 2. Press Instant Stop and then Reset after the cycle has been completed. XX-2 3. Place the cards in the read hopper in the following order: a. Fratar condensed program deck b. Card type 21 c. Card type 03 k. Press LOAD. 5. The program will load and then immediately begin operation. 6. After about 30 seconds, the message "TYPE IN NUMBER OF APPROXIMATIONS XXXXX PRESS RS KEY" will be typed. You must then type the five digit number of approximations (e.g. 0000^-) and hit the RS key. 7. Another message "TYPE IN NUMBER OF ZONES XXXXX" "PRESS RS KEY" will be typed. Type in the five digit number of zones which are in the trip tables (e.g. 0008^4-). Press the RS key and the data cards will be read. 8. The machine will now punch a type 22 card for each zone in the trip table. 9. Zero trip ends will not process correctly, jf there are zero trips ends for a zone in the trip table, instructions will be typed out on the typewriter. Follow the instructions to restart the processing. 10. The computer will continue for about 15 minutes after punching all of the type 22 card output. At this time the message "APPROXIMATION NUMBER 00001 COMPLETE" will be typed. If the option is specified, type 03 cards will be punched for this approximation and then the machine will continue. If the option has not been specified, the machine will continue the processing immediately. XX- 3 11. The processing described in step 10 will continue for each iteration as specified in step 6. Upon typing the message for the final iteration, the type 03 trip cards for that approximation will be punched. 12. After punching the trip cards, the type 23 cards will be punched. Following this, the message "PROCESSING COMPLETE" will signal the completion of the run. G. ERROR MESSAGES AND PROGRAM HALTS Typed Message Machine Action Explanation PROCESSING COMPLETE" Halts. End-of-Job. To process another set of data, place the required data cards (step 3') in the read hopper. Press reader start and console start and then follow the instructions beginning with step 6. "NO TRIP ENDS FROM Halts. Type in the present number of ZONE " trip ends as a five digit "TYPE IN PRESENT NO. number e.g. 00066. Then press OF TRIP ENDS XXXXX" the RS key to continue processing. "PRESS RS KEY" H. TIMING An 84 zone problem required 1 hour and 5 minutes to load and punch the trip tables for each approximation. I. METHOD The program first reads the type 21 growth factor cards. Then the type 03 trip table cards are read and the present trip ends are computed. The present trip ends are multiplied by the growth factor to obtain the number of desired (future) trip ends. If there are zero trip ends for a zone, a message XX-4 is printed out with instructions to follow in restarting the program. As the desired trip ends for a zone are calculated, the type 22 card is punched. After a type 22 card has been punched for each zone, the Fratar formula, the present trip ends and the growth factors are used in obtaining the nondirectional trips between each zone pair. At this time the program tests for the punch option. If the option has been specified, the trip cards are punched for each approximation. If this were the last approximation, the trip cards would be punched without regard to the sense switch option. After punching the trip cards for an intermediate approximation, new growth factors are computed by dividing the desired future trip ends by the calculated trip ends from the previous approximation. The formula is applied again and a new set of trips between zones is calculated. The punch option is checked and the entire process is repeated until the trip cards for the final approximation are punched. At this time, the type 23 cards are punched and the end-of-job message is typed. J. INPUT CARD TYPES - Card columns not specified must be blank. 1. Card Type 21 - Growth Factors Card Columns Content 1- 2 Card type, "21 3- 5 City code 9-11 Zone 18-22 Growth factor xx-5 2. Card type 03 - Output of Trip Table Builder Prog Card Columns Content 1- 2 Card type, "03" 3- 5 City code 6 Purpose code 9-11 Origin zone 12-14 Low destination zone 16-20 Volume 1 21-25 Volume 2 26-30 Volume 3 31-35 Volume 4 36-40 Volume 5 41-45 Volume 6 46-50 Volume 7 51-55 Volume 8 56-60 Volume 9 61-65 Volume 10 66-70 Volume 11 71-75 Volume 12 76-80 Identification (optional) K. OUTPUT CARD TYPES - Card columns not specified may be zero or blank. 1. Card type 22 - Present and Desired Trip ends Card Columns Content 1- 2 Card type, "22" XX-6 Card Columns Content 3- 5 City code 9-11 Origin zone 23-28 Present trip ends 40-45 Desired (future) trip ends 2. Card type 03 - Nondirectional Trip Volumes Same as input type 03 cards except that the volumes are non- directional. 3. Card type 23 - Calculated (future) trip ends Card Columns Content 1- 2 3- 5 9-H 40-45 L. PROGRAM DESCRIPTION Prepared in June 1965 by Clyde E. Sweet, Jr., Urban Planning Division, Bureau of Public Roads. Card type, "23" City code Origin zone Calculated (future) trip ends XX-7 U.S. DEPARTMENT OF COMMERCE BUREAU OF PUBLIC ROADS TRAFFIC ASSIGNMENT AND DISTRIBUTION FOR SMALL URBAN AREAS APPENDIX A - REFERENCES September 1965 (1) Traffic Assignment Manual , by the U.S. Department of Commerce, Bureau of Public Roads, Office of Planning, Urban Planning Division, Washington , D . C . , June 196k . (2) Calibrating and Testing a Gravity Model with a Small Computer , (IBM 1401) by U.S. Department of Commerce, Bureau of Public Roads, Office of Planning, Washington, D.C., October 1963; available from Superintendent of Documents, U.S. Government Printing Office, Washington, D.C. 20^02. (3) Manual of Procedures for Home Interview Traffic Study- -Revised Edition , by the U.S. Department of Commerce, Bureau of Public Roads, October 195^. (k) An Evaluation of Simplified Procedures for Determining Travel Patterns in a Small Urban Area , by Const ant ine Ben, Richard Bouchard, and Clyde E. Sweet, Jr., presented at k^rd. annual meeting of the Highway Research Board, January 196^, to be published. (5) Analysis of the Person Trip Data Voluntarily Submitted in a Dwelling -Unit Survey in Two Census Tracts in Youngstown, Ohio, Ohio State Highway Department, January 1963- (6) External Survey Manual , Niagara Frontier Transportation Study Manual; New York State Department of Public Works, Subdivision of Transportation Planning and Programming; Albany, New York, December 1962. (7) Truck and Taxi Survey Manual , Niagara Frontier Transportation Study Manual; New York State Department of Public Works, Subdivision of Transportation Planning and Programming; Albany, New York, December 1962. (8) Land Use Measurement Manual and Land Use Classification Manual , Niagara Frontier Transportation Study Manual; New York State Department of Public Works, Subdivision of Transportation Planning and Programming; Albany, New York, December 1962. (9) Contingency Check Manual, Niagara Frontier Transportation Study Manual, Number 110. M5; New York State Department of Public Works, Subdivision of Transportation Planning and Programming; Albany, New York, December 1962. (10) The Use of the Gravity Model for Forecasting Urban Travel-An Analysis and Critique, by Richard Bouchard, and Clyde E. Pyers, presented at l*3rd annual meeting of the Highway Research Board, January I96U, to be published. (11) Estimating and Forecasting Travel for Baltimore by Use of a Mathematical Model, by A. M. Voorhees and R. Morris, Highway Research Board Bulletin 22^, 1959, PP- 105 -ll 1 *. (12) Evaluation of Gravity Model Trip Distribution Procedures, by W. G. Hansen, Highway Research Board Bulletin 3hJ, 1962, pp. 67-76. (13) Nev Orleans Metropolitan Area Transportation Study, vol. 1, 1961, and vol. 2, 1962. (1*0 Integrating land Use and Traffic Forecasting, by C. F. Barnes, Highway Research Bulletin 297, 1961, pp. 1-13- (15) Travel Patterns in 50 Cities, by Frank B. Curran and Joseph T. Stegmaier, Highway Research Board Bulletin 203, 1958- (16) Future Highways and Urban Growth, by Wilbur Smith and Associates, New Haven, Connecticut, February I96TT PEf J N 'STATE UNIVERSITY LIBRARIES AD0QD712L 34134