Research Of Traffic State Space Model And Algorithm At Intersection

Traffic controller is an important component of urban traffic control system (UTC). It plays an essential role in improving the efficiency of urban roads and mitigating traffic congestions. Signal control algorithm is one of the key techniques of traffic controller.The concepts and methods of traditional signal control and traffic model are reviewed. A one-step prediction algorithm is proposed based on the state space model at intersection. The algorithm takes into account of the performance index and other basic restrictions of urban traffic. The phase composition, order and duration are computed accordingly. So it is different from the traditional algorithms which are based on fixed phase plans. Compared with the traditional fixed-time control, OSPA is more flexible in phase setting and can adapt to the variation of traffic conditions more easily. Simulation results show that the OSPA has a lower average parking delay time than the traditional fixed-time control and variable-phase algorithm control for various saturation rates when the traffic flow data is video captured traffic data.A Matlab? program for traffic simulation at intersections based on "psycho-physical" model, car following model and bicycle model is introduced. The program can simulate car stopping, queuing and turning at intersections reasonably and give delay time and numbers of stops. It can evaluate the performance of signal control algorithms when connected to the prototype traffic controller through serial port.