Traffic signals control system based on intelligent recommendation

Abstract

In order to contribute in facing the current challenging problems related to traffic congestion, we propose an intelligent traffic light control system called TIR-Light. Our system is based on the combination of recommendation systems and deep learning algorithms to maintain a fluid flow, prevent traffic jams and ensure a good quality of service for road users. TIR-Light predicts the hourly flow of each intersection and identifies the optimal timing planes by minimizing the waiting time and queue length. In addition, system efficiency and performance as well as its intelligent algorithms are evaluated through a simulation of the road network.