Problem Definition
This competition provides the traffic flow data, traffic scenario data (single intersection), the simulator and a basic signal plan to the participants. During the first stage of the competition, the participants are required to provide the traffic signal plan for that traffic scenario. During the second stage, the participants are required to submit the model that generates the signal plan. The final result is the travel time of vehicles based on the traffic signal plans.
As shown in the figure below, this competition will take a traffic scenario as input for a traffic simulator. The participants are required to provide the traffic signal plan for that traffic scenario. The final result is the travel time of vehicles based on the traffic signal plans.
Traffic signal control takes traffic situation as input. This input comes from the simulator, which requires road network and traffic flow as its input.
Road network describes the geometrical structure of the road network (roads, lanes, intersections). The candidate phases for each intersection are also included in the road network file.
Traffic flow data describes the planned route of a vehicle. For each vehicle, we have a starting time ts, and a traveling route consisting of each road segment that this vehicle will pass through.
Traffic signal controller generates traffic signal plan and the simulator takes that as input to run the traffic for the next second. For one intersection, the output of the signal plan is a number per second. This number indicates which phase this intersection will use for this second. In particular, a mandatory 5-second yellow phase (represented as 0) is applied if and only if when the phase is going to change, e.g., from 1 to 2.
For those vehicles which complete the trip from the starting road to the ending road, we define the timestamp it reaches the destination (enter the last road segment) as t_e. Then the travel time is defined as the time difference between its starting time t_s and ending time t_e and it is calculated by:
As for the rest of the vehicles which have not reached the destination when the system time is over at timestamp T_max, their travel time is calculated as:
Then for one test scenario, the metric Travel Time is calculated as the average travel time of all vehicles:
where n is the number of vehicles.
Obviously, the objective of this problem is to seek a better signal timing plan to minimize the average travel time T.
For each test scenario i, we have a baseline travel time b_i, then the score of the ith scenario is computed as the relative improvement over the baseline results:
The baseline result is based on the result from SOTL algorithm, which is provided in the sample code.
For n different scenarios, the overall score is:
