Dynamic, Integrated Model System: Sacramento-Area Application, Volume 2: Network Report (2014)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

26 A p p e n d i x B Create Working Folder For the integrated DynusT–FAST-TrIPs model, each executable file is assigned to a specific folder as shown in Figure B.1. Basi- cally, the data folder (given any desired name) includes all the main functions (executable files) of the model and the data set input files except those related to demand from the activity- based model. More specifically, the python master code, called DST_FT_module.exe is in this folder along with DynusT exe- cutable and FAST-TrIPs executable files (ft_Intermodal.exe, ft_Assignment.exe, and ft_Simulation.exe). In addition, two FAST-TrIPs python scripts (ft_BST.py and FAST_TrIPs.py) that call the main functions also need to be placed in this folder. A folder called VehicleDemandGen should be created in the main folder, in which all other python scripts are placed to generate demand files from the activity-based model. The two subfolders in VehicleDemandGen folder are assigned for the input files and output files. The main input file for the DynusT–FAST-TrIPs model is veh_sout.dbf, which is located in the Input Files subfolder, and all the demand-related files are placed in the same folder after generation in the bootstrapping step. Under the same token, DynusT output files need to be copied to the Output Files subfolder. Output Files folder is temporarily activated in terms of file transfer. The outputs (i.e., vehicle.dat and path.dat) generated by DaySim2DynusT-MAINprog-S.py in Vehicle- DemandGen are temporarily stored in the Output Files folder, then transferred to the working folder and used by DynusT during runs. Aside from hosting the demand file, the Input Files sub- folder also hosts information mapping files for DynusT and DaySim attributes as well as defined corridors or paths of interest and several DynusT network files. The mapping files are of importance for the following reasons: they define DynusT to DaySim zone, node, and parcel equivalents and coordinate information is used for distance calculation. The defined corridor data provide for easy access information for the purpose of post-processing simulation data. Lastly, the DynusT origin and destination network files will provide information for trips. As mentioned previously, the python files in the Vehicle- DemandGen folder read and translate DaySim trip infor- mation and prepare it as DynusT input files. To execute such processes, intermediate mapping files (found in Input Files subfolder) for DynusT and DaySim attributes must also be developed. In essence, this exercise ensures trips will start, originate, and terminate at the DaySim designated times and locations but within DynusT. Within the DynusT simulation a person making a trip will be assigned a vehicle; once this trip reaches its destination the assigned vehicle will exit the simulation. A person making multiple trips within the 24-h simulation period will be assigned into several vehicles; only one of those vehicles will be simulated at once. Table B.1 illustrates typical files (not an exhaustive list) for DynusT and FAST-TrIPs. Rancho Cordova network example Rancho Cordova, California, is an eastern suburb of Sacra- mento, California. The DynusT Rancho Cordova network was created in late August in 2010 and is a “subarea network cut” from the regional Sacramento model. The DynusT network is shown in Table B.2. Network This network is composed of 121 zones, 452 nodes, and 864 links. Like the downtown model, the Rancho Cordova network must first be cleaned/fine-tuned in order to bring it from its original state, a macroscopic model network, to mesoscopic model compatibility. For Rancho Cordova area, the transit network consists of 163 stops, 170 bi-directional transfer links, and 205 trips for Working with Data Set

27 Table B.1. Input File List in Working Folder DynusT Network Data (network.dat) Movement Data (movement.dat) Vehicle Generation Data (origin.dat) Vehicle Exit Data (destination.dat) Signal Data (control.dat) Simulation and Assignment Configuration Data (system.dat) Model Scenario Data (scenario.dat) Work Zone Data (workzone.dat) Incident Data (incident.dat) Vehicle Trip Schedule Data (vehicle.dat) Vehicle Route Data (path.dat) Advanced Parameter Data (parameter.dat) FAST-TrIPs ft_input_stops.dat ft_input_routes.dat ft_input_trips.dat ft_input_stopsTimes.dat ft_input_accessLinks.dat ft_input_transfers.dat ft_input_transitVehicles.dat ft_input_park-n-rides.dat ft_demand_auto.dat ft_demand_transit.dat ft_passengers.dat DynusT–FAST-TrIPs networks.dat AltTime_transit.dat AltTime_interModal.dat travelerVehicleMapping.dat travelerVehIntermodal.dat TransitRouteSchedule.dat TransitDwellTime.dat Figure B.1. Folder configuration. Table B.2. Input File List in “VehicleDemandGen–Input Files” Node-LatLong.geo Parcel-LatLongZone.geo parcelMOD.txt DSTtoDYSzones.txt PathInput.txt pathInputNAME.txt xy.dat network.dat origin.dat destination.dat veh_sout.dbf five bus routes. The Sacramento light rail transit (Gold line) is excluded in this data set. The transit data are shown in Figure B.3. Demand Exogenous trip tables (airport, external, thru auto and truck, and 2- and 3-axle commercial vehicles) from the travel demand model were converted to DynusT origin–destination (OD) demand table input. The conversion produced two temporal OD files, demand.dat and demand_truck.dat, with defined DynusT formats (explained in demand.dat description memo). The following table summarizes the generated demands.

28 Similarly, trip information from the activity-based model (DaySim) is converted, generating a vehicle.dat file with defined DynusT formats (explained in vehicle.dat description memo). The following table summarizes the generated demands. As previously mentioned and observed, the DaySim and DynusT values are different, due to the filtering trips modes that were not solely auto (e.g., “shared ride” or “drive alone”), such as “drive-transit-walk,” “bike,” and “walk,” along with other modes. Transit Route The operating transit system in Rancho Cordova is a part of Sacramento Regional Transit. There are five bus routes and a light rail line that cover Rancho Cordova area and the con- nection between other parts of Sacramento regional area (i.e., some of the routes are extended out of Rancho Cordova). To construct the transit network, General Transit Feed Specifica- tion (GTFS) data of Sacramento Regional Transit are used, and the required data are extracted for Rancho Cordova. The data were accessed in November 2010 through http://www .gtfs-data-exchange.com, and 2011 transit service was used for the network preparation. The transit routes serving Rancho Cordova are bus routes 21, 28, 72, 74, and 75, and a part of light-rail transit route number 507 (Gold line). By mapping the GTFS routes using GIS and comparing the data with the auto network and the Figure B.2. Rancho Cordova transportation network in DynusT. Figure B.3. Rancho Cordova transit network. Table B.3. Rancho Cordova Exogenous Demand Statistics File Name Trips Total Trips demand.dat • airport: 12,262 • thru auto: 220,489 • external: 202,886 435,637 demand_truck.dat • 2- and 3-axle: 83,388 • thru truck: 5,878 89,266

29 area boundary, the transit routes were cut to keep the segments inside the area only. In the same way, transit stops inside the area were selected. Since the transit network is used both in DynusT (for transit vehicle simulation in congested traffic network), and FAST-TrIPs (for assignment and simulation of transit passengers), transit routes had to be coded into the DynusT network. In other words, the path for each route (including its directions and variations) were found and stops were mapped into DynusT links, so that DynusT could read the transit network, generate buses like a type of vehicle, and simulate them in the model. This process was done in a semiautomatic way, so that paths were found by mapping GTFS shape files (containing geographic information about routes and stops) into the auto network and finding the best path that can represent the transit route. Obviously, the results were not perfect and many checks and corrections were done manually to accommodate transit routes in the DTA network (e.g., adding some missing links in the DynusT model on where transit vehicles travel). At the same time, the GTFS data, includ- ing stops, routes, trips, and stop-times (detailed schedule of transit vehicle trips) were prepared in appropriate formats to be read by FAST-TrIPs. Table B.4. Rancho Cordova Trip Roster Demand Statistics File Name Total Trips DaySim trip roster 140,099 vehicle.dat 117,795