casim

This project provides CA pedestrian simulation that is based on the article

Crociani, L. and Lämmel, G. (2016). Multidestination Pedestrian Flows in Equilibrium: A Cellular Automaton-Based Approach. Computer-Aided Civil and Infrastructure Engineering, doi: 10.1111/mice.12209

Getting started

Build

The easiest way to build casim is using Maven in the commandline. A jar file with all dependencies can be generated by invoking:

mvn test
mvn assembly:single

This will generate a jar file with all dependencies, which will be located in the subdirectory target

Run as stand-alone application

Even though casim is designed to be coupled to other transport simulation frameworks such as MATSim or SUMO it can also be used as a stand-alone application. An entry point to setup your own stand-alone scenario is the class:

org.cacrowd.casim.run.MyTestRunner

Some sample environment layouts can be found in the resources folder.

Run as hybrid simulation

casim offers an interface to be coupled with other simulation frameworks using protobuf over GRPC. In this setup casim acts as a server where other client simulation frameworks can connect to. A demo client-server setup is provided by the class:

org.cacrowd.casim.hybridsim.run.MyHybridSimTestRunner

Run hybrid simulation server in DIY docker container

casim server can also be run in a docker container. You can build and run the docker container by invoking

docker build --rm -t casim .
docker run -p 9000:9000 casim

Run hybrid simulation server in provided docker container

Docker builds are available via docker hub. To run casim in a provided docker container enter

docker pull grgrlmml/casim
docker run -p 9000:9000 grgrlmml/casim

Repeating experiments discussed in the CACAIE article

Thus far the Nash equilibrium and marginal social cost based approach experiments (see, sec 4.1 in the article) are available.

First you have to build the package

mvn test
mvn assembly:single

Then you can run the Nash equilibrium experiments by invoking

java -cp casim-0.1-SNAPSHOT-jar-with-dependencies.jar org.cacrowd.casim.matsimintegration.DaganzoExperimentRunner Nash <bottleneck width>

and the marginal social cost based approach experiments by invoking

java -cp casim-0.1-SNAPSHOT-jar-with-dependencies.jar org.cacrowd.casim.matsimintegration.DaganzoExperimentRunner MSCB <bottleneck width>

where must be one of {0.4, 0.8, 1.2}. Unlike in the article where 100 learning iterations with 2000 agents have been performed, the current setup runs the 100 iteration with only 500 agents. You can change this values in the code (see org.cacrowd.casim.matsimintegration.scenarios.DaganzoScenarioGenerator).

Per default every 10th iteration of the simulation runs will be visualized in real time in the visualizer gui. You can accelerate/decelerate the visualization with the +/- keys on your keyboard.

CACrowd/casim