多车型随机动态交通分配和信号优化组合

Combined multiple vehicle types stochastic dynamic traffic assignment and signal optimization

研究了多车型随机动态交通分配和信号优化的组合问题,提出了一种考虑多车型下的随机型动态交通分配和交通信号优化的组合模型。模型采用广义双层规划来表示,上层是信号优化控制,进行交通信号的优化配置;下层是考虑多车型下的随机型动态交通分配,进行交通网络流的配置。同时,模型中采用具有物理排队的多车型动态网络模型,从而考虑了饱和路网中物理排队对网络条件的影响以及不同车型间的相互作用。采用遗传算法对模型进行了优化求解。结果表明,在考虑多车型随机用户路径选择行为的基础上使整个网络的总行程时间达到了最小,由此获得最佳的信号控制策略和相应的优化路径流,实现了交通信号的优化配置和交通流优化。

The combined problems of multiple vehicle types stochastic dynamic traffic assignment and traffic signal optimization were studied.A combined model on multiple vehicle types stochastic dynamic traffic assignment and traffic signal optimization in saturated networks was presented.The model is formulated as a generalized bi-level programming problem.The upper level optimizes signal settings,while the lower level represents a multiple vehicle types stochastic dynamic traffic assignment,which sets network traffic flow.Meanwhile,the model applied a physical-queue dynamic network model with multiple vehicle types,which considers the impact of physical-queue on the network conditions in saturated networks and interactions between multiple types sharing the same road infrastructure.The solution of traffic signal and traffic assignment was optimized by using genetic algorithm.Simulation result shows that the total travel time can be minimized in considering multiple vehicle types stochastic user route choice behavior,the optimal signal control strategies and the corresponding optimal path flow are achieved,and the optimization settings of traffic signal and optimization of network traffic flow can be realized.