Optimized Control Of Traffic Signal For Urban Grade Intersection

Traffic problems with the mark of congestion have been the main obstacle for cities developing. People have tried many ways to improve the status. The direct mean is to build more overbridges to enlarge the traffic capacity, but that solution needs substantive fund and is limited by the actual city space; those are the difficulties to put it in practice. So, improving the level of traffic control and management and make best uses of the existing equipments are the effective methods. With the development of computer science, AI and detecting techniques, people integrate various techniques to control the urban signals to relieve the traffic congestion, make the traffic flow moving smoothly and minimize the loss.At present, the controlling method can be divided by their relationship of space as Isolated Intersection Control, Artery Intersection Control and Area-wide Control. It is also can be divided by the theory of control as Fixed-time Control, Actuation Control and Optimized Control. Any two from the different sorts can be fitted together freely. Isolated Intersection Control is to manage the signals of urban grade intersection as one unattached unit, whose aim is to minimize the delay at the end of its cycle. It is the main method of our country now; Artery Intersection Control is to manage some grade intersection at one time, whose aim is to make the vehicle flow missing or less-missing red signals while overpassing the artery. Area-wide Control is to manage many grade intersections of some region whose aim is to let some data such as stop times, traveling time, and oil wastage smallest. Fixed-time Control makes the equipments running as plan which is made by historical traffic data. Actuation Control is a method on-line control according to traffic capacity; it has no fixed cycle and green ratio. It sets vehicle detector at stop-line, and the cycle and green ratio changed with the detected data. Optimized Control comes into being with the development of computer and communications. It coordinates the cycle,green ratio and offset by the detected data to make the effect of traffic control best.In this paper, a different optimized control project for urban grade intersection is brought forward. At first, more advanced digital-radar detector is chosen instead of vehicle loop detector. Vehicle loop detector can be damaged by heavy vehicle and amending road, so it will take wrong count of vehicle. Digital radar detector can write down traffic data for 8 roadways at one time, and reports accurate and real-time traffic capacity, average rapidity and appropriate rate of roadway. In Isolated Intersection Control, Fuzzy illation is used to get cycle fitting for current traffic status, and adaptive genetic algorithm is used to assign green ratio for each phase. Adaptive chosen strategy is invented to weed week units at the initial stages. Sigmoid Function and evolution parameter are imported into adaptive crossover and mutation probability, so that the probability can change with the evolution process and go to a stable value at the end. In Artery Intersection Control, the method in Isolated Intersection Control is used at first to get cycle and green ratio of each intersection. And then, the real velocity which is detected on-line is changed into average velocity which is the basis of computing offset. With the result, artery intersection can be coordinated control.Conclusion after emulation mode:For Isolated Intersection Control, the effect of the algorithm of fuzzy-adaptive genetic in this paper is much better than the conventional method which is made by experience In four-phase control and three-phase control the length of cycle is reduced by 26.7% and 23.02% respectively, and the number of lagged vehicles is reduced by 38.92% and 38.84% under the same traffic circumstance; and compared with the algorithm of fixed cycle-simple genetic in four-phase control and three-phase control, the number of lagged vehicles is reduced by 33.33% and 32.77%. For Artery Intersection Control, the number of lagged vehicles of the algorithm invented in this paper is lower by 28.31% than the conventional method. So, the new method is correctand effective.Chief contributions:1,Fuzzy theory and adaptive genetic algorithm are integrated. Fuzzy control does not need accurate mathematic model, and it is robust and good at anti-jamming. So it is suitable for traffic that is dubious, imprecise, nonlinear pure lagged system. In this paper, proper cycle is fuzzy deducted by detected data. Genetic algorithm is a computational model of creature evolution. As a new kind of whole optimized searching algorithm, it is simple, all-purpose, robust, parallel and wielded usage. In this paper, improved adaptive genetic algorithm is invented to assign green ratio to each phase with the aim of smallest number of lagged vehicle at the end of the cycle.2,In the improved adaptive genetic algorithm, adaptive chosen strategy is imported. In the chosen stage of evolutional process, the unit with low fitness will be discarded and unit with fitness which is higher than a certain value can be selected. Its purpose is to enhance the whole fitness of population to select better unit into mating pool. The certain value is called threshold and it is not fixed, it can change with evolution.3,In the improved adaptive genetic algorithm, Sigmoid Function and evolutionary parameter are imported into adaptive crossover and mutation probability. Sigmoid function can make the algorithm avoid stagnant development under some circumstance and can save patterns of better units at the latter of evolutionary process. The function of evolutionary parameter is to make the probability go with the evolution and at last, it go to a stable value which avoids making not constringency or lengthening the constringent process.