基于自适应差分进化的干线交通信号协调控制

摘要

In order to improve the efficiency and stability of the coordination and control on urban arterial traffic signal of traditional intelligent optimization algorithms, and to avoid flaws such as low precision solution and local optima, a dynamic self-adaptive differential evolution algorithm, named p-ADE, will be examined. In the p-ADE, a new mutation strategy and a self-adaptive parameter adjustment technique are proposed to balance local and global searches for the improvement on convergence accuracy and speed. Experimental outcome illustrates the p-ADE algorithm outperformed several state-of-the-art optimization algorithms. In addition, innovative coordination control methods are capable of offering better traffic signal patterns that can reduce the average delay of traffic flow. This new technology will ultimately help improve the traffic capacity of urban trunk road traffic.%为克服现有基于传统智能优化算法的城市干线交通信号协调控制方法求解精度低、易陷入局部最优等缺陷,将改进后的动态自适应差分进化算法p-ADE应用于城市干线双向交通信号的协调优化控制,通过优化干线交叉路口相位差减小交通流平均延误.p-ADE在标准差分进化算法基础上提出了新变异策略和参数动态自适应调整策略,有效平衡算法的局部搜索与全局搜索能力.通过与基于多种群免疫算法等协调优化控制方法对比,实验结果表明,p-ADE在收敛精度、速度和鲁棒性上相比较于多种先进智能优化算法均具有明显优势,可以为交通干线系统提供更优的相位差,有效减少干线直行交通流的平均延误,提高城市主干道交通通行能力.