华南理工大学学报(自然科学版) ›› 2025, Vol. 53 ›› Issue (6): 91-103.doi: 10.12141/j.issn.1000-565X.240440

• 智慧交通系统 • 上一篇    下一篇

环线电动公交车辆调度与司机排班联合优化

胡宝雨1  齐月1  贾佃精2  程国柱1   

  1. 1.东北林业大学,土木与交通学院,黑龙江 哈尔滨 150040;
    2. 辽宁省交通运输事务服务中心,辽宁 沈阳 110003




  • 出版日期:2025-06-25 发布日期:2024-12-06

Joint Optimization of Circular Line Electric Bus Vehicle Scheduling and Driver Scheduling

HU Baoyu  QI Yue  JIA Dianjing  CHENG Guozhu   

  1. 1. Northeast Forestry University, College of Civil Engineering and Transportation, Harbin 150040, Hrilongjiang, China;

    2. Liaoning Transportation affairs Service center, Shenyang 110003, Liaoning, China

  • Online:2025-06-25 Published:2024-12-06

摘要: 为了解决环线电动公交车辆和司机任务分配不均衡的问题,本文提出了一种联合优化调度模型,该模型主要通过顺、逆时针方向互相调配车辆和司机来提高整体利用率。在给定环形线路和人车不固定的情况下,综合考虑车辆行驶里程、工作量、充电桩数量、车辆充电时间、司机工作时间、休息时间等约束,以公交企业总运营成本最小化和时刻表总调整最小化为目标制定有序充电管理计划和车辆、司机调度方案。在求解方面,首先将混合整数非线性规划模型通过线性转换转化成线性规划模型,并使用CPLEX求解器得到调度方案;其次采用多目标粒子群算法(MOPSO)和基于ε约束处理机制的改进多目标粒子群算法(ε-MOPSO)分别求解调度方案,并通过网格法确保外部档案集的收敛性和均匀性。最后以北京市环线公交200路(内、外环)为例进行验证,并对比分析CPLEX求解器、传统多目标粒子群算法(MOPSO)以及本文提出的基于ε约束处理机制的改进多目标粒子群算法(ε-MOPSO)的计算结果。结果证实了改进算法的有效性,且优化后的调度方案分别将车辆数从28降低到23,共减少17.86%;司机数从28降低到25,共减少10.71%。降低了车队规模和司机数量,从而降低了企业总运营成本;时刻表平均每个发车时刻调整4.13分钟,发车更均匀保证了乘客的需求。提升了公共交通的运营效率,具有重要的实际意义。

关键词: 城市交通, 联合调度, 多目标粒子群算法, 环线电动公交, ε约束处理

Abstract:

In order to address the issue of imbalanced task distribution between electric bus vehicles and drivers in the loop line, a joint optimal scheduling model is presented in this paper. It mainly allocates vehicles and drivers in the clockwise and counterclockwise directions to enhance the overall utilization rate. Given the circular route and the non-fixed passengers and vehicles, an orderly charging management plan and the vehicle and driver scheduling scheme are formulated with the aim of minimizing the total operating cost and schedule adjustment by comprehensively considering the constraints of vehicle mileage, workload, the number of charging piles, vehicle charging time, driver working time and rest time. Regarding the solution, the mixed integer nonlinear programming model is transformed into a linear programming model through linear transformation, and the scheduling scheme is obtained using the CPLEX solver. Secondly, the multi-objective particle swarm algorithm (MOPSO) and the improved multi-objective particle swarm algorithm (ε-MOPSO) based on the ε constraint processing mechanism are respectively employed to solve the scheduling scheme, and the convergence and uniformity of the external file set are ensured through the grid method. Finally, this paper takes Bus 200 (inner and outer rings) of Beijing Ring Line as an example to verify and compare the calculation results of the CPLEX solver, the traditional multi-objective particle swarm optimization (MOPSO), and the improved multi-objective particle swarm optimization (ε-MOPSO) based on the ε constraint processing mechanism proposed herein. The results validate the efficacy of the enhanced algorithm. The optimized scheduling plan reduces the number of vehicles from 28 to 23, amounting to a total of 17.86%. The number of drivers drops from 28 to 25, with a total reduction of 10.71%. The reduced fleet size and the number of drivers thereby lead to a decrease in total operating costs. The timetable is adjusted by an average of 4.13 minutes per departure time, and the departures are more evenly distributed to guarantee the demand of passengers. It enhances the operational efficiency of public transport and holds significant practical significance.

Key words: urban traffic, joint scheduling, multi-objective particle swarm optimization, loop electric bus;ε constraint