Selfless Traffic Routing Plus (STR-SUMO-Plus)
This is the development version used in the paper:Enhancing Dynamic User Equilibrium Utilizing the Selfless Traffic Routing Model
High level over view From a high level this code: 1 - generates trips considering the start and end edges in main.py as well as edge_dependcies. 2 - uses duarouter to make simple shortes path trips using Djikstra's algorithm. 3 - uses duaiterate to take the output from the last step and split vehicles arcoss the network in order to equilize delay between vehicles 4 - takes output from the previous step and refines it prioritizing vehicles with lower deadlines to get higher quality route assignments
Routes are regarded by quality on 4 metrics: Estimated travel time, Number of vehicles sharing the edges encountered, The number of edges, The distance (This is in the self.route_information dict in STRBE_controller)
Guide to experiments
Currently the code should be set to run Round3 from the paper. If you run main you should end up with the exact same results as given in the paper.
Switch Experiments To switch to another round: simply go to main.py and change the variable "Round_name" to the name of any existing round in the configurations folder. So for example if I wanted to re-run Round1 of the experiments I would change Round_name = "Round3" ->Round_name = "Round1" This gives us access to important files namely:
This brings us to step 2 from the previous section so what we need to do here is take the file generated in step 3 (in this case static2_090.rou.xml) and drag it into the configurations folder. Next we need to head back to main.py and change the variable "file_name2" to static2_090.rou.xml.
At this point you should be able to run main.py again and recreate the experiments from the paper.
Creating New Experiments Creating a new experiment is simular to switching experiments. Begin by going to Round_name and changing the value to what you wish to name the experiement. In this case I will use the name "duel_35".
Up to step 2 is automated currently now all that is left to do is generate a k-shortest trips file (provided by duarouter).
To ensure that step3 is automated please uncomment line 386 "file_name2 = using_duaIterate(file_name,Round_name)" in order for duatierate to run and generate a file.
Unfortunately due to duaiterate wanting to be run in the same directory as main (I literally can't find a fix please help) you will need to take the output file str_sumo.str.xml and the map file (in this case the map file is test_vehicle_only_4corners.net.xml) as defined by the net-file value in myconfig.sumocfg in the configurations file and place them into a independent file for duaiterate to work on.
In our case you can use either /Route_setup/Dua_temp or /configurations/duaiterate grinding in order to get the relevent file you will need. python duaIterate.py -t str_sumo.rou.xml -n test_vehicle_only_4corners.net.xml -l 100 This command is from scripts.txt
Afterward duaiterate should run for 100 iterations with the last iteration being in the file 099, retrieve the file str_sumo_099.rou.xml and place it in configurations.
Ensure that file_name2 in main.py is set to str_sumo_099.rou.xml and you should be able to run your new experiment.
Notes: -Duaiterate set to 100 iterations typically took around 3 minutes to run on my computer. -The newest iteration of duaiterate will be used. -This Duaiterate implementation is exetremely rough, since we plan on using something other than duaiterate I am not going to try and fix it at this point.
All relevant files will be moved to the relevenat round folder as well as the configurations folder.
A rule of thumb here is the files that will be used are in configurations while the round folders are "saves" for their respective trips.