AN IMPROVED TRAFFIC LIGHT PHASE PARTITIONING METHOD BASE ON TRAFFIC FLOW

Abstract

As time goes on, the traffic flow also has a corresponding change and it may be asymmetry, which would greatly decrease the traffic efficiency. In order to improve the traffic efficiency, alleviate traffic pressure and save transport cost, we hope that the traffic light can play a similar role as a traffic police man to make real-time decisions according to the dynamic environment of traffic. So in this paper, we proposal a variable phase partitioning algorithm according to the real-time situation of traffic flow. To verify the validity of the algorithm, this paper build a multi-agent intelligent traffic simulation platform by using the microscopic traffic simulation tool SUMO and multi-agent intelligent simulation framework JADE. Our approach can effectively reduce the traffic delay which was proved by experiments in the paper.