Project Design and Organization of Autonomous Systems : Intelligent Traffic Light Control

Because of the increasing density of the traffic flow in urban areas there is a need for optimal performance of traffic lights. In this paper we will describe an existing approach of reinforcement learning applied to the optimization of traffic light configurations, and introduce a new approach. Our approach uses implicit cooperation between traffic lights, letting cars take into account the traffic situation of the road ahead. We will show that this approach outperforms the existing algorithm, using experiments performed with a traffic simulator.