The Research Of Traffic Signal Control Technology Based On Flow Prediction

Nowadays, with the high speed development of social economic, the penetration of car is rising quickly. Large number of cars leads to the pinch of the road resource. Road congestion has been one of the greatest challenges in big city. Therefore, it is significant to propose a kind of prediction model based on traffic flow and achieve the intelligent control of intersection signal. This model can deal with the traffic congestion problem and improve the traffic capacity.An intelligent traffic signal control module is designed in this article, which is using the prediction of traffic flow and intersection signal control as objects, and is decreasing the waiting time as purpose. The module is used to adjust the timing plans in hope of improving the congestion. The five major results are listed as follows:(1) This dissertation analyzes the principle, the research status and application of of the intelligent traffic control system and its key technology. Then, proposed overall structure design of traffic signal control system based on traffic prediction.(2) This dissertation Studies the method of traffic flow collection which is in the premise of background frame difference method for moving target detection method based on YUV color model. Taking a junction in Hangzhou as an example to verify the detection is efficiency and the error rate is low.(3) This dissertation also researches the mechanism of predictive control and the rolling optimization and real-time feedback, proposing the improved scheme of short-time traffic flow prediction and the model of traffic flow prediction based on fuzzy neural network (FNN).Then use ant colony optimization algorithm (ACO) and Particle swarm optimization algorithm (PSO) to improve the model.(4) This dissertation studies the traffic signal control model of the single cross multi-phase intersection. A genetic algorithm is used to update intelligent control parameters of the model, and get the optimal signal timing scheme. Thus, it can achieve coordinated control of signal of each phase in the intersection to ensure the optimal performance index of the system.(5) This dissertation verifies the traffic signal control model of the single cross multi-phase intersection based on flow prediction with simulation.