多车型混编公交车队发车时刻及车型联合优化

Joint Optimization of Departure Times and Vehicle Types for a Multi-type Bus Fleet

相较于传统燃油公交,电动公交系统在车辆续航里程、车型配置及运营经济性上差异较大,现有公交时刻表优化方法较少考虑时变客流需求、非均匀发车时刻及车型配置等多种因素的影响,难以兼顾乘客出行体验与公交运营成本。面向多车型混编的电动公交运营场景,针对现有两阶段优化模型在灵活性和最优性方面存在不足的问题,提出一种公交发车时刻和车型配置联合优化方法。该方法以最小化乘客平均等待时间和公交运营成本为目标,考虑线路所需车队规模的影响,并以此建立多目标公交时刻表优化模型。其次,采用改进遗传操作的NSGA-Ⅱ(Non-dominated Sorting Genetic Algorithm-Ⅱ)求解该模型,以得到该问题的Pareto最优解集,并从中筛选出代表服务优先、成本优先及均衡模式3种不同效益偏好的时刻表方案。最后,基于实际公交线路验证所提出的时刻表优化模型及其求解算法的有效性和适应性。结果表明:所提方法可得到符合客流需求变化规律且代表不同效益偏好的时刻表方案;与两阶段模型相比,该方法在均衡模式下的优化方案可减少乘客平均等待时间12.8%,降低线路运营成本5.7%,能够综合提升公交服务水平并改善运营经济性。

Electric bus systems significantly differ from traditional fuel-powered buses in terms of range limitations,vehicle-type configurations,and operating economics.However,existing bus timetabling methods seldom consider the impacts of time-dependent passenger demand,uneven departure times,and vehicle-type configuration,making it difficult to balance the passenger travel experience with the operating cost of the bus system.This study focuses on an operational scenario involving a mixed fleet of electric buses with multiple vehicle types.To address the shortcomings of the existing two-stage optimization models in terms of flexibility and optimality,an integrated optimization method is proposed for bus departure times and vehicle configurations.The aim is to minimize the average passenger waiting time and bus operating costs by considering the fleet size required for the route.To achieve this,a multi-objective bus-timetable optimization model was established.Furthermore,the nondominated sorting genetic algorithmⅡ(NSGA-Ⅱ)with enhanced genetic operations was employed to solve the model and obtain the Pareto optimal solution set for the problem.Subsequently,timetable schemes were selected from this set that represented three different utility preferences:service first,equilibrium mode,and cost first.Finally,the proposed timetable optimization model and its solving algorithm were verified based on a real-life bus route to demonstrate their effectiveness and applicability.The results indicate that the proposed approach can generate timetable schemes that conform to passenger demand flows and represent various utility preferences.Compared with the two-stage models,the optimized solution in the equilibrium mode using the proposed method reduces the average passenger waiting time by 12.8%and decreases the total operating cost by 5.7%.This comprehensive improvement can enhance the quality of bus services while improving operating economics.