基于动态规划的过饱和交叉口信号配时优化

Signal Timing Optimization of Oversaturated Intersections Based on Dynamic Programming

针对城市道路交叉口高峰期间的过饱和状态易导致溢流、死锁甚至路网瘫痪的问题,首先,通过上游交叉口放行需求优先请求和下游交叉口放行可行性确定相位请求优先级;其次,基于动态规划的方法,构建了以通过能力最大为优化目标、排队长度和绿灯时间等为约束条件、周期和有效绿灯时间为优化参数的过饱和交叉口信号配时优化模型;最后,运用熵权法确定相位优先级,据此采用动态规划模型实时负反馈确定各相位绿灯时间,得到最大通过能力下的信号配时方案。相关的案例分析结果显示,文章研究的模型可以使交叉口的整体通过能力提高3%,最大排队长度和总体延误分别降低24.3%和25.1%,交叉口整体运行服务水平由F级改善为E级。

Aiming at the problem of overflow,deadlock and even road network paralysis caused by oversaturation during the peak period of an intersection,firstly,the priority request of upstream intersection release demand and the feasibility of downstream intersection release were used to determine the priority of phase request.Secondly,based on the dynamic programming theory,while using maximum capacity as the control objective,using the queue length and the actual green time as the objective constraints,and using the cycle length and the effective green time as the optimization parameters,a signal optimization model for an oversaturated intersection was established.Finally,the entropy weight method was used to determine the phase priority.Moreover,the dynamic programming model real-time negative feedback was used to determine the green time of each phase,and the signal timing scheme under the maximum capacity was obtained.The results of relevant case analysis show that the model can improve the overall capacity of the intersection by 3%,reduce the maximum queue length and overall delay by 24.3%and 25.1%respectively,and improve the overall operation service level of the intersection from level F to level E.