Research On Traffic Signal Control Strategies For Artery In Urban Road

With the increasingly serious congestion problem, more and more people focus on the signal optimization at intersection. In this paper, the signal decision process is modified to ease the burden of the artery. The first adaptive control strategy is innovatively designed for optimizing signal. And then, the adaptive control strategy is extended for restraining overflow in the downstream intersection. The innovations and contributions are summarized as follows:With the platoon delay time model, the membership functions are designed for the different arriving types of the platoon. The delay time of platoon arriving at the downstream intersection in the arbitrary point-in-time will be computed by the comprehensive model with membership functions. Considering that the local optimum value will be chosen as the global optimum value by the hill climbing algorithm, the search mechanism of outer loop is used in the algorithm to add the global searching capability. The algorithm performance is measured by the testing function. The effectiveness of the control strategy is verified and the comparisons between the fixed signal control and real-time control strategy are shown in the simulation.The token transition theory is proposed to enrich the Petri net theory. The conditions of transitions are illustrated with the token transition and supervision arc. With the theory, the Petri net model is designed for the intersections in the artery. In the designing process, the additional phases and waiting zoon are considered to increase the authenticity of the model. The adaptive control strategy is innovatively designed with the Petri net model. This is the first time that the feedback control strategy is included in the traffic signal control. The resistance to the disturbance of traffic and the variation of parameters are reinforced in the control strategy of intersection system. The stability conditions are deduced by the Lyapunov function. The least squares support vector machine is used in the control strategy to predict the traffic flow. In the simulation, the data are gathered from the real artery system and the effectiveness of the Petri net model and control strategy are verified.The control strategy is improved to restrain the overflow happening at the downstream intersection in road network, and the Petri net model of the intersection is designed. The phase switching part is added in the signal module and the overflow restriction part is also shown in the lane module. Using the improved Petri net model of the intersection as the reference model in the adaptive control strategy, the length of queuing vehicles are kept in the anticipated length with little fluctuation. In the simulation, the ability of the overflow restriction is tested and the control strategy of restraining overflow is analyzed for the upstream intersection. In this paper, the data used in the simulation are collected from the real intersection. The time of vehicles passing the intersection and the traffic flow in different periods of time are analyze from the authentic data to improve the applicability of the paper and to add the effectiveness of the control strategies.