基于改进差分进化算法的交叉口信号配时优化[1]

李玉贞1，王世豪2

1．上海电科智能系统股份有限公司，上海 200063

2．河南警察学院，郑州 450046

【摘要】快速增长的机动车数量与有限的道路资源之间的矛盾日益突出。交叉口作为道路的重要组成部分，交通信号配时的合理与否往往制约着道路的通行效率。目前，常用的固定信号配时方案无法根据实时交通数据流而动态调整，降低了道路的运行效率。针对该问题，构建基于车辆总延误时间的目标函数，并提出一种改进差分进化算法（IDE）对其进行求解。该算法在个体变异阶段引入两种变异策略，从而实现全局探索能力到局部开发能力的转变。同时，利用正弦函数自动调整每代所需的控制参数。利用8个测试函数对IDE的优化性能进行评估，试验结果表明，IDE算法能够显著提高DE算法的收敛精度和稳定性。进一步将IDE算法应用于交叉口配时问题，仿真结果表明提出的IDE算法是一种求解交叉口配时问题的有效方法，能够降低车辆延误时间，提高通行效率。

【关键词】交叉口信号配时；延误时间；差分进化；变异策略

Timing Optimization of Intersection Traffic Signal Based on an Improved Differential Evolution Algorithm

Li Yuzhen1,Wang Shihao2

1．Shanghai SEARI Intelligent System Co.，Ltd.，Shanghai 200063

2．Henan Police College，Zhengzhou 450046

Abstract: The contradiction between the rapid growth of the number of motor vehicles and the limited road resources has become increasingly prominent. Intersection is an important part of the road, while the traffic signal timing often restricts the traffic efficiency of the road. At present, the commonly used fixed signal timing scheme cannot be adjusted dynamically according to the real-time traffic data stream, thus reducing the operation efficiency of the road. To solve this problem, an objective function based on total delay time is constructed, and an improved differential evolution algorithm （IDE） is proposed to deal with it. The IDE introduces two mutation strategies in the individual mutation stage, and thus realizes the transformation from global exploration ability to local exploitation ability. At the same time, the sine function is used to automatically adjust the control parameters required by each generation. Eight test functions are utilized to evaluate the performance of the proposed IDE, and the experimental results verify that IDE can dramatically enhance the con vergence accuracy and stability. Furthermore, the IDE is applied to the intersection signal timing, and the simulation results indicate that it is an effective method, and is able to reduce the vehicles delay time and improve the traffic efficiency.

Key words：intersection signal timing;delay time;differential evolution;mutation strategy