地铁乘客限流控制与列车运行图协同优化方法研究

Collaborative optimization method of passenger flow control and train timetable for metro systems

随着客流需求的快速增长,高峰时段客流需求的过度饱和给地铁系统运营安全和乘客高质量服务带来了巨大挑战。为了提高地铁运行效率,减少旅客出行延误,本文为客流过饱和条件下地铁网络运营提出了一种有效方法——客流控制策略与列车运行图协同优化。以乘客平均等待时间最小化为目标,综合考虑行车条件约束、站台容量约束及列车容量约束,建立了一个混合整数非线性规划模型,并设计了一种改进的差分进化算法,来寻找地铁网络限流控制与列车运行图协同优化问题的高质量解。最后,以北京地铁部分线路为例,将无优化、仅限流优化、仅优化运行图和协同优化四种策略的结果进行对比,验证了所提协同优化方法的有效性。结果表明:限流控制与运行图的协同优化策略最能提升运营效率,所提差分进化算法在计算效率和解的质量方面均优于遗传算法。协同优化策略下乘客的平均等待时间为6.02 min,与无优化相比所获得的限流控制与列车运行图协同优化策略可以使乘客平均等待时间减少17.65%,乘客平均出行时间减少4.42%。

With the drastic increase in travel demands,oversaturation during peak hours brings difficulties to the safety of metro operations and high-quality passenger service.A method of optimizing passenger flow control strategies and train timetables is proposed to improve system operation efficiency and reduce travel delays.Aiming to minimize the average waiting time for passengers,a mixed-integer nonlinear programming(MINLP) model is established by considering the constraints of train operation,and platform and train capacities.In addition,an improved differential evolution(DE) algorithm is designed to obtain high-quality solutions for the collaborative optimization problem.Finally,a case study of some Beijing metro lines is conducted to demonstrate the effectiveness of the proposed approaches.The comparison of the four plans—no optimization,flow-control-only optimization,train timetable-only optimization,and collaborative optimization—shows that the collaborative optimization plan can increase operational efficiency the most.The proposed DE algorithm outperforms the genetic algorithm in terms of computational efficiency and solution quality.The average passenger waiting time under the collaborative optimization plan is 6.02 min.Compared with the no-optimization plan,the obtained plan of collaboratively optimizing flow control strategies and the train timetable can reduce the average passenger waiting time by 17.65% and the average passenger travel time by 4.42%.