Step 3: Adding standalone and Feeder DRT using eqasim - Hussein-Mahfouz/matsim-drt GitHub Wiki

Adding Standalone DRT

I am following the instruction in the eqasim library for DRT. See the documentation here. I will go through the steps of adding a basic DRT service (not a feeder), and then go through the feeder. This is because the feeder builds on the basic implementation.

Step 1: Create DRT vehicle fleet (or multiple)

  • CreateDRTVehicles is used to create vehicles and add them at initial positions in the network

  • Implementation

    • a: Create a wrapper class to run it (see here)
    • b: Use a bash script to run it multiple times with different configurations (if you want different fleets in different areas). See example here

  • Notes on configuration 

    • Check the arguments you can pass here, and a description of the parameters here. some are missing

  • Notes on parameters

    • Shapefile polygon should be one contiguous polygon. It cannot be disconnected polygons
    • You can use serviceBeginTime and ServiceEndTime to create multiple fleets that operate in the same area at different times of day. For example, if the service area has 40 vehicles from 8am - 11am, and then 20 vehicles from 11am - 2pm, this can be configured by creating a different drt xml file for each. You would only change the time parameters and the vehiclesNumber parameter

Step 2: Adapt config for Drt

  • AdaptConfigForDrt is used to edit the config so that it works with the eqasim DRT extension

  • Implementation

    • Create wrapper class around AdaptConfigForDrt (see here
    • Run it using a bash script (see here

Step 3: Add to Mode Choice Module

Step 4: Run the simulation

  • Create LeedsDrtModule (same as corsica). This sets up the necessary components for DRT simulation, including utility estimators, cost models, and predictors
    • Add LeedsDrtModeAvailability to LeedsDrtModule
  • Create RunDMCSimulationDrt to run the simulation

Adding Feeder DRT

Use the same DRT fleet created above

Step 2: Adapt config for Feeder Drt

  • AdaptConfigForFeederDrt takes a config created by AdaptConfigForDrt, and edits it to work with FeederDrt.

  • Implementation

    • Create wrapper class around AdaptConfigForFeederDrt
    • Run it using a bash script

Step 3: Update ModeAvailability

In LeedsModeChoiceModule

  • Add the FeederDrtModeAvailabilityWrapper to the LeedsModeChoiceModule as a wrapper around the pre-existing ModeAvailability (see here

  • Give the mode availability a name (see here)

  • bind it to the FeederDrtModeAvailabilityWrapper class (see here)

Step 4: Utility Estimation

Option A: To use the DefaultFeederDrtUtilityEstimator class

Step 1: Edit the LeedsModeChoiceModule

  • Give the utility estimator a name (see here)
  • bind it to the DefaultFeederDrtUtilityEstimator class (see here)

Step 2: Edit LeedsDrtModule

  • bind utility estimator for feederDRT to the specified estimator (see here)

Step 3: Edit the config

  • edit eqasim::estimator for all feeder DRTs to DefaultFeederDrtUtilityEstimator (see here)

Option B: Create your own FeederDrtUtilityEstimatror class

  • Create my own LeedsPtDrtUtilityEstimator
    • Use both pt and drt predictors to get trip attributes
    • Calculate Utility based on how much each mode is used in a trip (similar to utility based on bus + rail)
  • Question: Will the router return trips that use both PT and DRT? Which part of the config allows this combination to happen?
    • Is it multimodeFeederDrt::accessEgressModeName? (see here)
    • If that is the case, do I need to specify this mode name (drtA / drtB) when writing my utility estimator? Or can I just use "drt"?