Design And Development Of ITS Based Intelligent Traffic Signal Control System

Under the existing road conditions,intelligent traffic signal control system is one of the most effective methods to improve the operating efficiency of traffic facilities,alleviate urban traffic jams,reduce environmental pollution and reduce traffic accidents.In this paper,the intelligent traffic signal control system is designed based on ITS framework concept.The proposed system consists of four core modules: Terminal Perception,Decision Analysis,Parameter Setting and Real-time Monitoring.The main work and contributions of this paper include:1.Based on the core idea of the ITS system framework,the abstract service content is embodied as the final product plan through the structured design method,which fully proves the practicability of the ITS design concept for the complex urban traffic system construction.2.A Terminal Perception module including data acquisition and data preprocessing is researched and developed.Firstly,this paper developed a data acquisition system based on MMW radar sensor and its network structure.Then,in order to improve the accuracy and structure of the data,this paper proposed a data preprocessing method based on Kalman filter algorithm and matrix correction algorithm.The actual road experiment results verify the effectiveness and feasibility of the terminal perception module.3.A real-time signal control method based on Deep Reinforcement Learning is proposed.In this paper,TD3 algorithm which is optimized based on DDPG is applied to traffic signal control for the first time.By using the simulation environment established by the intersection of Wenyi Road and Fengtan Road,the optimization effect of TD3 and DDPG algorithm is compared,which proves that TD3 algorithm has better optimization performance.4.The software platform of intelligent traffic signal control system is researched and developed.Using web development technology,the product design scheme proposed in this paper is fully implemented under the SpringMVC framework.At present,the four functional modules of the system have been basically realized,which can monitor the real-time road condition information,manage the terminal equipment and set their parameters,and recommend the scheme of signal duration for users for reference.