Research On Intelligent Control System Of Intersection For Hybrid Traffic Flow

With the rapid development of economy and the continuous increasing number of vehicle, the urban-development is greatly limited by transportation problems in our country. Information technology, sensor technology, communication technology, automatic control technology, operational research, image process technology, computer network and artificial intelligence are used into the whole control-traffic system by Intelligent transportation system(ITS), and ITS builds a real-time , accurate and efficient transit general system in wide field. The Traffic Control System is a complicated system, it has some factors including people, vehicles and roads. For avoiding traffic jam, we need get a way to reduce traffic volume at hot hours, and augment traffic ability. ITS is a good way to solve the traffic problem besides building roads and improving traffic environment. The intersections are the converge points and play an important role in urban traffic, while the traffic jams often happen at these intersections. So it is an important job to study the intersection sign control. In addition the traffic flow in China is hybrid mainly, and it is different from the other countries'traffic flow that makes up of vehicle flow. Therefore, studying artificial intelligence control of Chinese hybrid traffic flow about a traffic intersection has a great significance.The contents of the research as the following:(1) Hybrid traffic flow characteristics analysisAfter analyzed and summarized the researches of confused traffic flow characteristics at home and abroad, this paper summarizes and analyzes the hybrid traffic flow basic principle and characteristics and studies the interference of the hybrid traffic flow at a traffic intersection. The conclusion is the basis of Research on Method of Neural-fuzzy control under hybrid traffic flow of a traffic intersection.(2) Method of the BP Neural Network in Prediction of Traffic VolumeThe prediction of traffic volume can offer important data for the intelligent control of intersection and is one of the prerequisites of artificial intelligence control for traffic intersections. In all kinds of predictions, BP Neural Network is universally used in prediction of traffic Volume and has some success results. In order to make the predictions of BP Neural Network more accurately, the method of prediction of traffic volume is based on the subsection learning of double—layers BP Neural Network and the study uses the improved method to predict the traffic volume of Jingshi Road Jinan City, then compares the results which made by subsection-learning method and made by common method. Using subsection-learning method, the average relative tolerance is decreased by 2.52%. The improved BP neural network can be used for prediction of traffic volume.(3)Research on Method of Neural-fuzzy control under hybrid traffic flow of a traffic intersectionBasing upon the research on the theory of prediction to hybrid Traffic Volume, a new Hybrid Traffic Flow Fuzzy Control System Model of Intersection is made by analyzing the influence on Non-motorized vehicles to Hybrid Traffic Flow to Intersection Control system. This model's inputs are pressure degree and transform degree, and its output is the green light delay time which is happening meanwhile.MATLAB is utilized to simulate Hybrid Traffic Flow Fuzzy Control System of Intersection according to the data of the intersection Traffic Volume of Jingshi Road and Lishan Road. The result shows that it can reduce the average queue length effectively and raise the Traffic Capacity at intersection.