Enhancing Traffic Flow Efficiency through an Innovative Decentralized Traffic Control based on Traffic Bottlenecks

There are two main folders: data and code.

Inside the code folder, there are four independant code parts:

1. net_data_builder.py - analyze the SUMO network file and creates a network analysis json file input_network_data.json, that is used later to build the trees.

2. tripper.py - creates the trips file: vehicles.trips.xml for each simulation, with vehicles origin, destination and start time.
   It can use a random OD or a known probabilty.

3. run-multi.py - the main python projects, which uses SUMO's TraCI package to run multiple SUMO simulation according to the configuration.
   It uses the product of the two previous code parts.

4. figures_calculations.R - uses the results of the simulations to create the figures.

Inside the data folder, there are four experiments folders with the 20 runs of each experiment smulation per each traffic-light control method.

There is also an excel file: results.xlsx, which summarize all the results together.