基于改进遗传算法的平面交叉口信号优化控制法

本文对城市交通中单交叉口信号动态优化控制技术进行了深入研究,在此基础上设计了一种应用于单交叉口的智能信号控制优化算法,并在遗传算法的选择算子中对竞争法进行了改进,加入了希尔排序策略,将基本遗传算法改进成了一种新的基于二进制编码的遗传算法. 且计算机模拟复杂度较高的四相位交通控制仿真对比实验取得了良好的效果. 实验结果证明,遗传算法可以较好地应用到交通控制领域,且改进式遗传算法在中、重度交通需求的情况下依然能在很短的计算时间内使控制周期内路口的总延误和排队车辆数明显减少.

Isolated Intersection Control Based on Improved Genetic Algorithm

Single intersection control is an efficient tool to release traffic congestion. We proposed intelligent algorithms which can be used in single-intersection control optimization. In the algorithm, the competition law for the selection operator of genetic algorithms was improved with the Hill sorting strategy, and the basic genetic algorithm was changed into an improved new binary genetic algorithm. Simulation results from a 4-phase interaction control strategy show that genetic algorithm is applicable to traffic control. Genetic algorithm improved in the paper reduced delay and queuing length significantly within short time under high traffic demand. This new method of signal control gives a new idea for the signal control and provides more theoretical basis in the area of the optimization of signal control for the future.