3 Model Steps
nitabhave edited this page Sep 5, 2023
·
1 revision
This table contains the name and description of each model step along with which input files are used and which output files are created or edited.
| Sub-step (rsc file) | Brief Explanation | Input Files | Output Files |
|---|---|---|---|
| Initial Processing | |||
| Create Initial Output Files (01 - CreateScenario.rsc) | Scenario folder structure and geographic files are produced. | master network, master route system, master TAZ layer | scenario network, routes, TAZs |
| Network Calculators (02 - NetworkCalculations.rsc) | Initial network calculations (capacity and speed); network and SE data error checks; area type and smoothed area type calculated; CC speeds, bus speeds and other attributes set; highway and transit networks created. | scenario hwy line layer, scenario transit route system, various factor tables | prepared line layers, networks and skims |
| Accessibility (03 - Accessibility.rsc) | Calculates accessibility measures: Gini-Simpson Diversity Index, Walkability Score, Insersection terms. Calculates attractions used for walkability model, percent of a TAZ near bus stops, and TNC wait times. | base attraction rates, accessibility parameters, all streets layer | modified SE data, accessibiity skims |
| Population Synthesis (04 - PopulationSynthesis.rsc) | Calculates popsyn marginals from ACS data using disaggregate curves for HHsize, income and workers. Runs TransCAD population synthesizer. Matches HH marginals and attempts to match person marginals using PUMS for seed data. Determines HH auto ownership. | PUMS seed data, ACS marginals, SE data | synthesized HHs, synthesized population |
| Generation Models | |||
| Home-based Productions (05 - ResidentProductions.rsc) | Household and person data are modified to create seniors, children, worker, accessibilities, etc. The estimated production factors are applied to get zonal trip productions by trip type. Separately, calibration factors are applied that provide transparency on the original and the final calibrated values. | HH file, person file, SE data and production rates | resident trip productions by time of day within scenario_se.bin |
| Airport (Airport.rsc) | Simple airport model that generates and distributes airport trips. Mode choice occurs within the feedback loop and uses N_HB_O_Long probabilities. | directionality, model and tod (3 csv files) | airport_auto_trips, airport_pa_trips, airport_transit_trips (x 4 time periods = 12 matrices) |
| University (University.rsc) | Separate model that generates and distributes university trips by 6 different trip purposes and 5 different modes. | modes (5), university attraction rates, directionality, gravity, production rates, tod, trip rates (11 csv files) | university_trips (x 4 time periods = 4 matrices) |
| Externals (External Models.rsc) | Model that generates and distributes EE, IE and EI trips. | ee - tod and seed ieei - directionality, gravity, model, tod, transit (7 csv files) | ie_pa_trips.mtx, ie_od_trips.mtx |
| Commercial Vehicles (Commercial Vehicles.rsc) | Applies a basic QRFM approach with Ps = As and GM trip distribution. | generation, gravity by tod, tod (6 csv files) | cv_gravity (x 4 time periods = 4 matrices) |
| NonMotorized Models | |||
| NonMotorized Choice (06 - NonMotorized.rsc) | Vehicles per adult, income per person, and age variables are calculated. The binary choice model is applied by looping over each trip type and spliting trips into motorized and non-motorized. Non-motorized indicators are added to the person table, giving person level data for NM trips. | household and person data, accessibilities, walk accessibility attractions, walk skims, coefficient files | daily aggregated NM trips (bin) NHB NM trips by trip purpose (7 bin files) nm_gravity matrix |
| NM Distribution (06 - NonMotorized.rsc) | Trips are aggregated to TAZs and a gravity model is applied using walk accessibility attractions and walk skims. | n/a | n/a |
| NM Time-of-Day (06 - NonMotorized.rsc) | The trip table is split into NM trips by TOD using the same TOD factors. | n/a | n/a |
| Time of Day | |||
| Aggregation to Zones (07 - Aggregation.rsc) | Up to this point the models are disaggregate (by persons and/or HHs). At this step, data is aggregated by TAZ and added to the sedata output file. Market segments get collapsed except for work trips. | Disaggregate person and HH data | scenario_se.bin |
| Time of Day Split (08 - Time of Day.rsc) | TOD factors get applied to resident HB models. The remainder of the model steps will feedback by TOD. | TOD factors and resident trip productions | resident trip productions by TOD in scenario_se.bin |
| Skimming | |||
| Roadway Skims (10 - Skimming.rsc) | A skim matrix for the highway network is created which includes zone-to-zone congested travel time, distance, and toll cost. Also because TOD happens later, directionality is applied here to create skims that are a weighted average of the PA and AP travel time. | highway network, TAZ file, SE table, DirectionFactorsSkims tabel. | output |
| Transit Skims (10 - Skimming.rsc) | Skims for each transit mode and access mode by TOD are created. | transit route layer, transit mode table, access mode, transit network (tnw) | output |
| Home-based Choice Models | |||
| Parking Probabilities (11 - Parking.rsc) | Auto travelers parking location and mode (walk or transit) from parking lot to final destination is determined; only applied to parking districts (CBD and universities with limited, paid parking). Results of the parking model influence both mode choice and destination choice. | SE data, walk skim | output |
| Mode Probabilities (12 - Mode Choice.rsc) | Aggregate mode choice probabilities between zonal pairs are calculated. | SE data, synthesized household table, transit mode, access modes, roadway and transit skims, trip production rates | output |
| Destination Probabilities (13 - Destination Choice.rsc) | Jobs are split into high- and low-earning and then destination choice probabilities are calculated. | SE table, Args.ResDCAttrRates, Args.ResTODFactors, Args.ResDCSizeCoeffs, skim files, shadow price | output |
| Application of Probabilities (13 - Destination Choice.rsc) | A destination choice matrix by trip purpose and TOD is created. | SE table, destination choice probability matrix, mode choice probability matrix | output_matrices |
| Non-Home-Based Models | |||
| NHB Generation by Mode (14 - NHB Models.rsc) | NHB trips based on HB trip attraction ends are generated. | HB trip matrices, SE table, TOD factor, NHB generation parameters | nhb_dir + “/generation.bin” |
| NHB Destination Choice (14 - NHB Models.rsc) | The destination choice models for NHB purposes are run. | SE table, nhb_dir + “/generation.bin”, intraClusterMtx = skims_folder + “IntraCluster.mtx”, NHB DC coefficients | NHBTripsForDC.bin, NHB model files, NHB probability files, total NHB trips |
| Special Market DC & MC | |||
| Airport MC (Airport.rsc) | Airport trips are generated and distributed. | SE table, highway skim files, airport model file, highway network, airport_tod factor | Airport trips production in SE table, airport_pa_trips.mtx |
| University DC & MC (University.rsc) | University trips are generated and distributed by mode and TOD. | SE table, university production rates, attraction rates, tod, gravity, highway skims, mode share probability, directionality factor | Productions and attractions added to SE table, university_**.mtx |
| Commercial Vehicles DC (Commercial Vehicles.rsc) | CV trips are generated and distributed. | SE table, CV trip rate, highway skim | cv trip generation in SE table |
| Roadway Assignment | |||
| Create Assignment Matrices (16 - Create Assignment Matrices.rsc) | Four TOD matrices for HB, NHB, Airport and University trips are generated. Auto person trips from resident matrices are categorized as SOV1, HOV2, HOV3 using occupancy/ shares factors and are then converted to vehicle trips assigned using directionality factors. Trip conservation snapshots are taken along the way to see if trips are lost. | directionality_factors.csv; other_shares_hb.csv | trip_conservation.csv |
| Pre Assignment (17 - Roadway Assignment.rsc) | Network check to see if HOV lanes exist. If present, HOV lanes will be excluded from SOV and truck assignment. | vot_params.csv | n/a |
| AM/MD/PM/NT Assignment (17 - Roadway Assignment.rsc) | Assigns trips in four TOD matrices to the roadway network in parallel order with feedback loops for each TOD period. | scenario_links, od_veh_trips_**.mtx, vot_params.csv, net_TOD_mode.net | scenario_links and net_TOD_mode.net |
| Post Assignment (17 - Roadway Assignment.rsc) | Updates congested travel time on the route network, bus speeds, and %RMSE for skims used to achieve convergence. | scenario_links, skim_sov/hov_TOD.mtx | scenario_links, skim_sov/hov_TOD.mtx |
| Convergence | |||
| Convergence (17 - Roadway Assignment.rsc) | Calculates %RMSE for each zonal pair on SOV/HOV TOD matrices used in assignment (transit links and low demand ij pairs are ignored). Once the %RMSE for CongTime for all pairs, for each purpose, show minimal residual difference from the previous CongTime, the model converges. | scenario_links, skim_sov/hov_TOD.mtx | avg_skim_TOD_Purp_hov/sov.mtx |
| Hourly Roadway Assignment | |||
| Peak Hour Roadway Assignment (17 - Roadway Assignment.rsc) | Assigns peak hour volumes of the PM period (pkhr factor = .39) to one hour of capacity | scenario_links, od_veh_trips_PM.mtx | pmpk_hr_assn.bin, od_veh_trips_PM_PKHR.mtx |
| Transit Assignment | |||
| Create Transit Matrices (18 - Transit Assignment.rsc) | Creates matrices including airport, NHB, university and external transit trips. | n/a | trip_matrices |
| Run Transit Assignment (18 - Transit Assignment.rsc) | Assigns combined transit matrix files to transit network and route system. | mtx_file, net_file | FlowTable, OnOff, TransitLinkFlows, WalkFlow |
| Summaries | |||
| Maps (19 - Summaries.rsc) | Creates maps that show count difference; speed by TOD; VOC by TOD for LOS D & LOS E. | several | output_summaries |
| Calibration Reports (19 - Summaries.rsc) | Creates summary reports that show model results. | several | output_summaries |