Multi-phase Urban Traffic Signal Real-time Control with Multi-objective Discrete Differential Evolution

Abstract

A real-time and on-line intelligent urban traffic signal control approach based on forecasting technology is proposed for the requirements of real-time traffic signal control. Considering the multiple performance indicators, the multi-phase multi-objective point-control model and line-control model are respectively constructed. By combining the characteristics of real-time urban traffic signal control, a multi-objective discrete differential evolution is introduced to solve the multi-objective optimization models. Meanwhile, a real-time traffic flow sequences generating program is designed to simulate the actual urban traffic flow which is subject to the Poisson distribution. Furthermore, a traffic signal on-line control and performance evaluation platform is developed. A set of simulations are conducted by this platform. The performance of the proposed approach is compared with that of F-B approach which is state-of-the-art. Experimental results show that the proposed approach can obtain the better traffic signal configuration and have a good real-time performance.