Distributed Traffic Lights Control based on Swarm Intelligence

Abstract

In this study, we aim to smooth the traffic flow by controlling the traffic signal appropriately. this paper propose two decentralized multi-agent signal control systems. The first is a split and offset control method in which each agent controls traffic light based on traffic information obtained from itself and neighboring agents using spring model. The second is a wide-area offset control method that forms organizations by neighboring agents cooperating with each other based on local traffic information, and makes green wave. In addition, it is indispensable for these methods to install sensors at each intersection, but it is not necessarily such a rich environment. Therefore, this paper presents a traffic information interpolation system using Ant Colony Optimization, which is one of the swarm intelligence algorithms.