Intersection Intelligent Control Based On Fuzzy Control And Genetic Algorithm

With the rapid development of urbanization, the problem of city transportation is becoming bigger and bigger to each city in China. More and more phenomenon indicates that the taffic jams usually take place in the city intersections. This dissertation takes a single intersection as research object because China's transportation control systems are predominant by single-intersection control.In order to decrease the average length of queues in a single intersection, this dissertation designed severial self-adaptive fuzzy controllers by using the intelligent control technologies which combined with Fuzzy Control(FC) and Genetic Algorithm(GA).The main contents of this dissertation include:Firstly, the main control parameters, the performance indexes, the meaning of single-intersection control and its traditonal methods were explained detailedly.Secondly, a multiphase classical FC which aimed on decreasing the average length of queues was introduced.Thirdly, a multiphase two-stage FC was proposed to overcome some drawbacks of classical FC. Not only the fuzzy rules can be acquired easily, but also its numbers can be decreased effectively. The phasic order also can be optimized at the same time.Fourthly, a multiphase FC based on GA was designed to overcome some drawbacks of classical FC. This part includes: the design of a new GA which inquired a learning process which can effectively overcome the drawbacks of"early maturing"and bad globle convergency of classical GA; The optimization of fuzzy rules'adjusting numbers by using the new GA.Fifthly, all the controllers were simulated by MATLAB and comparison research was fufilled under different transportation enverionments. The results showed the superiority of two-stage FC and FC based on new GA.In the end, summarized the whole work and pointed out the directions of further research.