Coordinate Optimization Control Of Traffic Signal Based On Combined Algorithm Of Simulated Annealing And Elite Co-evolutionary Algorithm

With the improvement of the level of city, the number of motor vehicles has increased a lot,which significantly aggravates the traffic congestion, and the method of traditional traffic signalcontrol has not been able to achieve good results. Therefore, intelligent traffic signal control hasbecome the main means. Using evolutionary algorithm for optimal control of traffic signal is aneffective way, but the convergence precision and the convergence speed still needs to be improved.co-evolutionary algorithm is a new intelligent algorithm, because of its highly parallelism, it has cuta striking figure in the field of intelligent traffic signal control. This paper improves theco-evolutionary algorithm, then uses it for traffic signal control and achieves good results.Based on the elite strategy and the concept of co-evolutionary and the simulated annealingmethod, this paper put forward a new algorithm, which is the elite co-evolutionary algorithm basedon the simulated annealing method (SACEA), and then it is used in the field of traffic signal control.SACEA have three populations: elite population, common population and stochastic population.Elite individuals form teams and exchange information with other team members through thecooperating or leading operation. SACEA introduce the stochastic population to evolution toimprove diversity of population, at the same time, the stochastic individual and the elite individualusing fast simulated annealing method to realize the purpose of cooperation. Through all above, thealgorithm get a better global optimization performance and a faster convergence speed. Evaluationexamples have shown its validity.This paper had studied the urban single-point intersection traffic flow characteristics, based onthis, two dynamic traffic flow models were set up. These two models respectively take theminimum average delay time and average parking rate as the control objective. On this basis,SACEA were used for the optimization and simulation of these two models. Compared with theexiting optimization algorithm, the results show that: using SACEA to optimize can moresignificantly reduce the single intersection average delay time and average parking rate.Based on analyzing the theories of urban line coordination control, this paper selects a greenwave system as an example, and selects the average total time(average travel time plus totalaverage delay times)as the control target, and then makes a timing optimization of multiintersection by SACEA. Compared with the exiting algorithm, the results show that: vehiclesthrough the line need less time by using SACEA to optimize the multi intersection signal timing.