基于IC卡数据的公交行车计划优化研究

发布时间：2018-10-11 19:38

【摘要】：公交行车计划的编制是将各车次分配至各公交车辆的过程。因此对其的优化包括了决定车次的公交时刻表以及场站的公交车辆数。优化的目标是减少场站配车数并且解决目前存在的各公交车辆之间载客量、总运行时长不均衡等问题,为各公交车辆构建均衡合理的车次链。而行车计划的优化过程中必不可少的一项即为乘客的出行数据,目前常用的数据获取方式为人工调查,此方式在耗费大量人力财力的同时,精确度得不到保证,因此本文采用乘客出行的公交IC卡刷卡数据作为数据来源对公交行车计划进行优化。首先,基于公交IC卡数据判断乘客的上、下车站点。通过刷卡时间差临界值区分两个相邻的刷卡记录是否位于同站,在此基础上建立站点上车人数计算算法,并通过Pathon语言实现程序化的站点上车人数计算。此外,论文通过建立假设与集合的方式,将公交出行链法与站点吸引权重法相结合,通过乘客下一次上车站点推导乘客的下车站点,以求得各站点的下车人数;然后,引入断面客流量的计算方法,以建立期望载客量模型,该期望载客量模型考虑了公交运营成本以及乘客舒适性两方面的因素,确定出车辆合理的载客人数,即期望载客量值。为保证各发车车次均衡载客,选择载客量较大的前几个站点作为关键站点,构建关键站点累计载客量曲线,确定出各关键站点累计客流量曲线达到期望载客量值时所需要的发车时间,并选取最小的发车时间作为优化后的发车时间,以此类推,依次得到各时段优化后的发车时间;之后,需要将优化后的时刻表所对应的各车次安排给各公交车辆,因此建立了基于发车时刻表的场站逆差函数,并建立不同场站间空驶车次的插入算法,在时刻表优化不变的条件下,减小场站逆差函数高峰值,即减少配车数。在优化场站车辆数后,各车次按照FIFO并且车辆总运行时长均衡的规则进行分配,构建各车辆的车次链,得到均衡的行车计划。本文选取了广州市23路以及37路作为实例进行分析,证明本文提出的行车计划优化方案,不仅在满足客流需求的同时减少了场站所需的配车数,并且解决了平峰时段部分车辆过于拥挤而部分车辆载客不足、各公交车辆车次链总运行时长不均衡等问题,为各车辆分配了合理的车次链,形成一个完整的行车计划。
[Abstract]:The compilation of bus plan is the process of distributing each train number to each bus vehicle. Therefore, the optimization includes the bus schedule that determines the number of buses and the number of buses at the station. The aim of the optimization is to reduce the number of buses assigned to the station and to solve the existing problems such as the capacity of each bus and the imbalance of the total running time so as to build a balanced and reasonable train number chain for each bus. One of the essential items in the optimization of train planning is passenger travel data. At present, the commonly used data acquisition method is manual survey, which consumes a lot of human and financial resources, but the accuracy can not be guaranteed. Therefore, this paper uses the IC card data of passenger travel as the data source to optimize the bus plan. First of all, based on the bus IC card data to judge the passenger on, get off the station. According to the critical value of the time difference of credit card, two adjacent credit card records are located at the same station or not. On this basis, an algorithm for calculating the number of boarding passengers at the station is established, and the program calculation of the number of boarding stations is realized by Pathon language. In addition, the paper combines the bus trip chain method with the station attraction weight method by establishing the assumption and the set, and deduces the passenger's alighting station through the next boarding station, so as to find out the number of the passengers at each station. This paper introduces the calculation method of cross-section passenger flow to establish the expected passenger capacity model, which takes into account the two factors of bus operation cost and passenger comfort, and determines the reasonable number of passengers carried by the vehicle, that is, the expected passenger load value. In order to ensure the balanced passenger load of each train, select the first few stations with large passenger capacity as the key stations, and construct the cumulative load curve of the key stations. Determine the departure time required when the accumulated passenger flow curve of each key station reaches the expected capacity value, and select the minimum departure time as the optimized departure time, and so on, and then get the optimized departure time of each time period in turn; After that, it is necessary to arrange the trains corresponding to the optimized timetable to each bus, so the deficit function of the station based on the departure schedule is established, and the algorithm of inserting empty train numbers between different stations is established. Under the condition that the schedule is optimized, the peak value of the deficit function of the field station is reduced, that is to say, the number of cars allocated is reduced. After optimizing the number of vehicles in the station, each vehicle is allocated according to the rules of FIFO and the equilibrium of the total running time of the vehicle, the train number chain of each vehicle is constructed, and the balanced driving plan is obtained. In this paper, the 23 and 37 roads in Guangzhou are selected as examples. It is proved that the train planning optimization scheme proposed in this paper can not only meet the demand of passenger flow, but also reduce the number of bus distribution required by the station at the same time. It also solves the problems of overcrowded part of vehicles and insufficient carrying capacity of some vehicles in the flat peak period, and the total running time of each bus vehicle is not balanced, which allocates a reasonable train number chain for each vehicle and forms a complete train plan.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：U491.17

【参考文献】