Multi Phase Intelligent Fuzzy Simulation Research And Implementation

Traffic simulation is important for traffic management. It can improve the efficiency of traffic control. This dissertation mainly focuses on the traffic simulation using configuration technique. The main work of the dissertation includes:Based on the support vector machine (SVM), a new method is proposed to predict the short-time traffic flow volume. The structure of the SVM model is dependent on the historical traffic data in the last hour. The experimental results show that this prediction method based on SVM is efficient. Especially during the period from 7:00 a.m. to 22:00 p.m., the mean squared error and absolute mean error of the prediction model are 6.0veh/5min, 5.1veh/5min, respectively.Based on the configuration technique, a multi-phase traffic simulation system of intersection is designed. According to the scale of project and programming convenience, Kingview 6.53 is chose as the programming platform. The simulation object is a typical crossing signalized intersection, and eight lanes from four entrances are simulated. In this developed simulation system, multi-phase control, fixed control and fuzzy control all can be actualized. Man-machine interface is friendly, and all kinds of simulation parameters can be adjusted on-line. The running of this traffic simulation system showed that it can provide an intellective support for department of traffic police.