交通运输工程与信息学报

传统城市轨道交通应急接驳公交蓄车点选址的研究对时间的处理过于简化,未充分体现乘客的时间需求,针对这一不足,引入时间容忍度概念。首先,根据出行计划是否受到影响及受影响程度,对系统内乘客进行分类,分析轨道交通中间站及端点站在乘客类别上的差异。而后,利用时间容忍度的调查数据,拟合满意度函数,并据此定义本文的时间惩罚成本函数。进而,以蓄车点数量最少和时间惩罚成本最低为目标,建立蓄车点选址-分配模型,研究乘客等待时间容忍度对选址的影响。最后,以轨道交通车站均为中间站的上海长宁区为例,选用Lingo11.0对ε-约束法处理后的模型进行求解及灵敏度分析,并提出蓄车点服务能力限制下的模型改进方法。案例结果表明,在时间容忍度为[5,10]min时,公交接驳服务无法令乘客满意;在时间容忍度为[20,25]min时,可最大程度减少蓄车点数量;考虑乘客等待时间容忍度后,蓄车点并非总是服务于与其最近的需求站,而同一需求站一般不会由多个蓄车点协同服务。

In general, traditional research on depot locations of emergency bridging buses for urban rail transit severely simplify the time while failing to reflect fully passenger time demands. To solve this problem, time tolerance is introduced in this study. First, passengers in the system are classified according to whether the trip plan was affected and to what extent, and differences in passenger categories between intermediate and terminal stations of urban rail transit are analyzed. Second, based on survey data concerning time tolerance, a satisfaction function is fitted and a time penalty cost function is defined. Then, taking a minimum number of depots and the lowest time penalty cost as the objective function, a model for allocating depot locations is established to investigate the effects of passenger waiting-time tolerance on location. Finally, Changning District, Shanghai(where urban rail transit stations are all intermediate stations) is taken as an example for numerical study of the time tolerance model. After the model is processed using the ε-constraint method, Lingo11.0 is selected for model solving and sensitivity analysis, and an improved method for the model under the limitation of depot service capacity is proposed. Results of the numerical example showed that when the time tolerance was set to [5, 10] min, the services of emergency bridging buses were unable to satisfy passengers. When the time tolerance was set to [20, 25] min, the number of depots could be minimized. Finally, when the waiting-time tolerance of passengers was considered, the depot did not always serve the nearest demand station, and a single demand station was generally not served by multiple depots.