A traffic light control method based on multi-agent deep reinforcement learning algorithm

Architecture of the proposed traffic light control method

In this section, the proposed multi-agent based ITLC algorithm will be presented. Two factors are considered by an agent who executes the proposed algorithm. One is the traffic condition of corresponding intersection. The other is traffic condition of neighbor intersections. As the vehicles that appear at the neighbor intersections may arrive at current intersection in the next moment, traffic condition of neighbor intersections should be considered as well. So two steps are executed by the proposed algorithm. Firstly, the agent will…