A Comprehensive Regional Traffic Coordination Control Strategy Integrated the Short-Term Traffic Flow Identification and Prediction

Abstract

Traffic congestion caused by the rapid growth of vehicles and unadaptable signal control has become a major constraint on traffic efficiency and travel experience. However, the existing studies have primarily focused on a particular aspect of traffic behaviors, without considering the systemic and interconnected nature of traffic congestion. Hence, a comprehensive strategy integrating monitoring, prediction, and coordinated control at regional intersections is proposed to improve traffic efficiency. First, an intelligent algorithm is designed to identify and predict traffic condition information. Then, the intersections with similar traffic behaviors and higher relevancy are divided into the same subzone, and a multiobjective optimization model is proposed to improve traffic capacity and green time utilization. Furthermore, the timing is modified according to the temporal and spatial characteristics of the oversaturation. Simulation studies in long-term and short-term traffic congestion scenarios are designed to verify the performance of the proposed strategy. Compared with the existing multitime timing method and a newly formed reinforcement learning control method, the results show that the average delay of the proposed strategy is 4.945% and 6.737% inferior to the outstanding strategies.