针对过饱和状态下的交通信号配时,以区域整体输出总流量最大化和各交叉口进口道总延误时间最小化为目标,构建了基于动态子区划分的交通信号区域协调控制双层规划模型（BP模型）.通过分析交叉口滞留排队车辆、进口道交通量以及相位相序对区域协调控制的影响,建立了交叉口相位差、有效绿灯时间和动态交通流量等协调控制变量的约束关系式.采用基于遗传–模拟退火算法的混合优化策略对模型进行求解,提高全局最优解的可靠性和计算效率.仿真结果表明,该模型与实际情况较为吻合.

 Aiming at the traffic signal timing in oversaturted traffic conditions,a bi-level programming（BP） model for the area coordinated control of traffic signals is presented based on the dynamic subarea division,with the maximum total throughput and the minimum delay time of all entrance lanes at each interaction as the objectives.Then,the effects of queue length,entrance traffic volume and phase sequence on the area coordinated control are analyzed,and the constraint formulas describing the relationships among the intersection phase difference,the effective green time and the dynamic traffic volume are established.Finally,the proposed model is solved by using the hybrid optimization strategy of genetic and simulated annealing algorithms,thus improving the reliability and computational effi-ciency of the global optimal solution.Simulated results indicate that the proposed model accords well with the practical situation.