地铁应急接驳公交派车场站选择与行车计划协同优化

Integrated Optimization of Dispatching Depot Choice and Scheduling for Emergency Bridging Bus in Response to Metro Disruption

地铁网络发生区间故障后,广泛采用应急接驳公交维持公共交通的服务保障能力。为提高地铁应急接驳公交运行效率,减少应急公交数量和乘客时间损失,以公交备选派车场站是否被选择、派车场站的公交停放能力以及应急接驳公交快慢车方案为决策变量,建立地铁应急接驳公交派车场站选择与行车计划协同优化模型,通过线性化处理以及ε-约束法,将模型转化为单目标线性规划模型。以某地铁线网为例,通过Gurobi求解模型的帕累托最优前沿,结果表明:相比无快车模式,所提方法在不增加公交数量的情况下,可增加总发车次数,使得乘客时间损失降低4.17%;应急公交数量和乘客时间损失具有负相关性,决策者可根据实际需求权衡两个目标;公交派车场站应尽可能地均匀分散选择;增加公交派车场站数量可降低乘客时间损失,但增加派车场站数量产生的边际效益逐渐降低,实际派车场站数量可根据实际情况酌情确定。

After an interval failure occurs in the metro network, emergency bridging buses are widely used to maintain the service guarantee capability of public transportation. To improve the operation efficiency of emergency bridging buses and reduce the total number of bridging buses and passengers′ time loss, an integrated optimization model of dispatching depot choice and scheduling for emergency bridging bus was proposed. The decision variables were whether the alternative dispatching depots were selected, the bus parking capacity of each depot and express/local plan for bridging buses under emergencies. The original model was transformed into a linear programming model with a single objective by linearization and ε-constraint method. Taking one metro network as an example, the Pareto optimal frontier of this model was determined by Gurobi solver. The results showed that the proposed method was able to increase the total number of departures and reduce the passengers′ time loss by 4.17% without increasing the number of buses compared with the no express bus method;there was a negative correlation between the total number of bridging buses and the passengers′ time loss, and decision-makers could realize a trade-off between these two objectives according to the needs;the chosen dispatching depots should be distributed as evenly as possible;increasing the number of bus dispatching depots could reduce passengers′ time loss, but the marginal benefit of increasing dispatching depots was gradually declining. Therefore, the actual number of dispatching depots can be determined according to the actual situation.