基于遗传算法的交通信号动态优化方法

针对典型的城市多车道双向交叉路口的交通流分布，以四相位信号控制为例，建立了以控制周期内路口的总延误车辆数最小为控制目标、以信号相位绿灯持续时间和信号周期时长为控制变量的交通信号动态配时模型。并用基于实数编码的遗传算法对信号周期和相位4绿信号时间等控制变量同时进行优化。为检验算法的优化效果，针对实际交叉路口高峰小时的实测交通流量数据，进行了大量次数的仿真计算，并对仿真结果进行了分析。

A Dynamic Optimization Method for Traffic Signal Timings Based on Genetic Algorithm

For a single two-way intersection with multi-lanes, a dynamic traffic flow model based on four-phase control is formulated, in which the total number of the delay vehicles over the control period is adopted as the control objective and the phase green times and cycle time as the control variables. A real-coded genetic algorithm (RGA) is used for simultaneously optimizing the green times and cycle time. For examining the validity of the optimization method, large amounts of simulation computing are executed based on the peak-hour traffic flow data of a practical intersection, and the simulation result was thoroughly analyzed.