Adaptive Fuzzy Traffic Signal Control Based On GA & Traffic Simulation Software Development

With the fast development of the urbanization and the rapid increment of the vehicles in the urban area in the recent twenty or thirty years, the urban intersections become more and more crowded, and they turn to be the bottle-neck of the urban mixed traffic controlling. Therefore, using the automotive control method to make the whole urban traffic network work smoothly and safely is most important, and is emphasized by more and more scholars from both inside and outside. The research works of this field is becoming the most essential part of the Intelligent Transportation System.Since Fuzzy Logic Control can integrate with expert's experiences and without precise mathematic model, it has come to be the most popular method in this field. However, as the fuzzy rules and the membership functions of the traditional Fuzzy Logic Controller are always given by human experiences beforehand and once decided never changing during the whole control period, the coming problem is that when the traffic condition changes too much, especially, the vehicles flow of different directions are discrepant greatly, the control effect becomes worse. So the Genetic Algorithm is used to real-timely optimize the FLC's fuzzy rules and membership functions to form self-adaptive FLC because the Genetic Algorithm has got the strong searching ability without knowing the mechanism of the object and only driven by the fitness function. At the same time, in order to simulate the different signal control schemes using in the big scale traffic network of urban area, and analysis the result and evaluate the different schemes, the author's team also developed the Simulation and Analysis System of Urban Mixed Traffic.This thesis presents the total project of designing the self-adaptive Fuzzy logic Controller using in both isolated intersection and in urban traffic networks and some content of the Simulation System's development. The main content of the thesis is as follows:1 . Summing up the background of the traffic signal control in ITS and the researchwork of optimal FLC used in the intersection signal control. Classifying the designation method of all kinds of FLCs', and explaining the designing details. 2. Discussing designation of GA optimizing respectively the fuzzy rules andmembership functions of the traditional FLC using in isolate intersection signal control based on the analysis of the traditional FLC's structure. Explaining the coding scheme, fitness function, GA operator, etc details. The effectiveness is proved by the simulation result of MATLAB.3. Aiming at the difference between the urban traffic network and the isolatedintersection, designing the double FLCs system, and real-timed optimizing the sequential membership functions of the inputs and output of the Secondary FLC. Explaining the coding scheme, fitness function, GA operator, etc details. The effectiveness is proved by the simulation result of MATLAB.4, Presenting the designation and implementation of Simulation and Analysis Systemof Urban Mixed Traffic which was developed by the author's team. Now the system was checked and accepted by Economic and Commerce Committee of P. R. China ([2003] No.216) (see also in the appendix II). The system includes GIS management module, traffic simulation system's parameter setting module, the other third side'straffic light controlling program module, microcosmic simulation module and simulation results analysis module, which can realize the on-line simulation and off-line analysis function. Based on this, the implementation of the microcosmic traffic objects in the simulation kernel and the designation and realization of the graohic-user-interface program of the system is emphasized in the later chapter. In the end, the thesis concludes with a summary and perspectives of the future research of the field.