【作者】 于剑；

【导师】 张星臣；

【作者基本信息】 北京交通大学 ， 交通运输规划与管理， 2015， 硕士

【摘要】 列车运行计划是城市轨道交通的核心技术文件,指挥列车的日常运行,主要包括列车开行方案和列车运行图两个部分。随着城轨路网规模的扩大,运营组织呈现网络化特征,难度大大增加,这使编制合理的列车运行计划成为值得研究的问题。本文主要研究网络化条件下城轨列车开行方案编制技术和列车运行图换乘协调优化技术两项内容,成果适用于连通/非连通型城轨系统,具体包括：(1)轨道交通开行方案编制方法总结在阅读大量相关文献的基础上,系统总结了轨道交通系统列车开行方案的编制方法,着重从编制思路、优化模型、求解方法三个方面阐述。编制思路上,把常见思路归纳为分阶段法和整体法两种;优化模型上,总结了相关文献中2类共15个优化目标,以及7类共26个约束条件；求解方法上,将主流方法分为数学方法和启发式方法两类,分别概述了每类方法的内容。(2)网络化条件下城市轨道交通列车开行方案编制技术固定停站方案(站站停)和编组方案(给定列车类型、编组长度),研究OD客流量已知条件下的列车交路方案和开行频率优化方法。综合考虑运输企业和旅客双方面收益和成本,建立了网络化条件下城市轨道交通列车开行方案优化模型,并提出混合蚁群算法和遗传算法的启发式求解算法,求解过程包含的核心内容有：全网多路径搜索、备选交路集搜索、大规模网络下基于交路方案的全网多路径客流分配、断面客流已知下的开行频率最优化及启发式交路方案搜索。(3)网络化条件下城市轨道交通列车运行图换乘协调优化技术研究给定运行图的换乘协调优化技术,通过调整各线路列车到发时刻,减少旅客在换乘站的换乘等待时间。首先建立单站、网络换乘协调优化模型,之后提出基于贪欲思想和运行线整体平移思路的优化方法,可降低整个网络上的旅客换乘等待成本,并且不会对正常运输秩序造成过大影响。(4)城市轨道交通运行计划编制平台利用C#语言开发了城市轨道交通运行计划编制平台(PubTransCAD-URT),包括文件系统、基础路网、开行方案、列车运行图、优化计算5个主要功能模块,可完成从基础物理路网绘制,到开行方案、运行图编制,再到自动优化计算的整个运行计划编制流程,是产品级成果。(5)算例分析分别以杭州地铁和北京地铁网形为基础,虚构连通网络算例2个,详细展示了基础客流路径搜索、备选交路集搜索、全网客流分配、开行频率人工及自动优化、网络运行图换乘协调优化等过程的参数、结果及运算效率,证明了模型与算法的有效性。更多还原

【Abstract】 Train operation plan is the core technology document of urban rail transit and commands the daily operation of trains. It consists of2main parts:train line plan and timetable. Along with the extension of the urban rail transit network, the operation and organization process of trains has become even harder due to the network scale, making it worthy to research the compiling technique of the train operation plan. This article mainly focuses on2aspects:train line plan compiling technique in urban rail transit network, timetable coordination optimization considering the passenger transfer waiting time. The proposed models and algorithms can be used in both connected and unconnected urban rail transit networks. Detailed contents are:(1) Summary of the train line plan compiling technique of rail transport systemI systematically summarized the line plan compiling methods of rail transport system based on a number of relevant literatures. The summary includes3aspects: compiling ideas, optimization model and solution methods. Aspect1elaborates2main compiling ideas as system split and non-split. Aspect2introduces2categories of optimization objectives (containing15different objectives) and7categories of constrains (containing26different constrains) in relevant literatures. Aspect3classifies the solution methods as mathematical methods and heuristic methods and outlines each one of them.(2) Train line plan compiling technique on the urban rail transit networkThis part tries to find out good train lines and their frequencies under the condition of a fixed stop scheme (trains stop at every station they pass) and marshaling scheme (train type and number of carriages are set beforehand), with OD data fully known. A line plan model that considers both the benefits and costs of urban rail transit company and passenger is constructed. The solution is heuristic, mixed with the ant colony algorithm and genetic algorithm. Core solution procedures include:multi-route searching, potential lines searching, line plan based multi-route passenger flow assignment in a larege-scale network, lines frequency optimization based on known section passenger flow data and heuristic line sets searching.(3) Timetable coordination optimization considering the passenger transfer waiting time This part focuses on the coordination optimization technique of a given timetable, aiming to decrease the total transfer waiting time by adjusting the arrival and departure time of trains. Models of one single station and the whole network are respectively constructed. The solution algorithm uses the greedy algorithm and the idea of moving all timetable lines together. The proposed approach decreases the transfer waiting cost in the whole network and has no great influence on daily operations.(4) Train operation plan compiling platform of urban rail transitAn urban rail transit operation plan compiling platform named PubTransCAD-URT is developed using C#. This platform consists of5main modules including file module, basic physical network module, line planning module, timetabling module and optimization module. It can help fulfilling the tasks from basic physical network drawing, to line planning and timetabling and auto optimization of line plan and timetable. I personally consider this platform a product-level achievement.(5) Case studyBased on the network of Hangzhou Metro and Beijing Subway,2virtual connected network cases are constructed, where the main steps of the algorithms are operated, including the basic passenger route searching, potential lines searching, whole-network passenger flow assignment, manual/auto line frequency optimization and timetable transfer coordination optimization. Parameters, results and computing times are given, which proves the effectiveness of the models and algorithms.更多还原