Abstract

Abstract:-Vehicular travel is increasing throughout the world, particularly in large urbanareas. Therefore the need arises for simulating and optimizing traffic controlalgorithms to better accommodate this increasing demand. In this paper we studythe simulation and optimization of traffic light controllers in a city andpresent an adaptive optimization algorithm based on reinforcement learning. Wehave implemented a traffic light simulator, Green Light District that allows usto experiment with different infrastructures and to compare different trafficlight controllers. Experimental results indicate that our adaptive trafficlight controllers outperform other fixed controllers on all studiedinfrastructures.

Keywords:Intelligent TrafficLight Control, Reinforcement Learning