【作者】 黄鉴；

【导师】 彭其渊；

【作者基本信息】 西南交通大学 ， 交通运输规划与管理， 2013， 博士

【摘要】 客运专线旅客列车开行方案是客运专线运输组织的重要组成部分,它涉及到铁路相关固定设备和移动设备的合理运用,如何制定一个合理可行的旅客列车开行方案,是铁路旅客运输面临的亟需解决的核心问题。尤其是我国已将客运专线建设纳入国家重要发展战略规划,客运专线的大规模修建与持续投入运营,对于推进我国城镇化和现代化,统筹城乡区域协调发展,满足旅客出行需求具有十分重要的现实意义。为了配合客运专线的建设与运营,需要对列车开行方案的编制与优化问题进行及时研究,完善相关理论与方法。客运专线列车开行方案优化问题涉及的影响因素、优化环节众多,其中客流预测是列车开行方案制定的基础,而列车开行方案制定的优劣反过来也会影响客运专线客流预测结果,两者互为因果关系,因此,将两者结合起来进行综合优化研究更具合理性。本论文在参考国内外相关研究成果的基础上,研究客运专线客流预测及列车开行方案相关问题,通过建立相关问题的优化模型,设计了各问题的有效求解算法。论文的主要研究工作如下：1.总结了目前常用的客流预测方法,指出诱增客流预测相关理论研究存在尚不够系统的问题,对诱增客流预测方法进行了重点研究,提出了重力模型的改进策略,结合重力模型及生长曲线模型构建了诱增客流量预测的综合模型,以满足不同时期客运专线诱增客流的预测需要,并实现了成绵乐客运专线主要区域间诱增客流量的预测。2.旅客换乘网络是实现列车客流分配的关键,为了方便旅客乘车方案的表达及广义费用计算,提出了一种新型混合换乘网络,并设计了该网络的最短路算法。以新型混合换乘网络为基础,分别建立了用户平衡条件下、弹性需求条件下和容量限制条件下的列车客流分配模型,并设计了相关算法。3.分析了与列车开行方案密切相关的几个主要问题,包括客运专线运输组织模式,列车跨线运行的经济距离以及客运专线的运营收益。明确了我国高速客运专线适合的运输组织模式,列车跨线运行经济距离的计算方法,客运专线运输部门运营成本及运营收入的主要构成等。明确了列车开行方案的优化内容、开行方案优化问题的边界条件、开行方案的优化原则及优化思路,从客运专线列车的停站模式出发,建立了列车停站方案优化模型,模型以最小化旅客出行广义费用及剩余客运周转量为目标,兼顾了旅客的经济利益和社会效益,然后以列车停站优化模型为基础,引入最大化客运专线运营收益目标和最小化列车运行路径广义阻抗目标,增加列车开行方案相关约束条件,建立了列车开行方案优化模型,并对模型进行了分析。4.总结了智能优化算法研究概况,提出了基于群体多样性保持的和声搜索(DMHS)算法,该算法从音调微调概率动态调整方法及和声记忆库更新策略两个方面对基本和声搜索算法进行了改进,并利用TSP问题对DMHS算法的有效性进行了验证,结果表明改进后的算法不易陷入局优,优化性能显著提高。然后结合客运专线列车停站设置原则,设计了基于DMHS的列车停站优化算法,在此基础上,将DMHS算法嵌入模拟退火算法,设计了适合求解列车开行方案问题的混合智能优化算法,以西成客运专线列车开行方案优化问题为例,对混合优化算法的有效性进行了验证。5.研究了基于客流动态调整的列车开行方案综合优化问题,在分析客流预测模型的基础上,以最大化客运专线运输收益和最小化列车运行路径广义阻抗为目标,建立了客流预测与开行方案综合优化模型,并设计了综合优化模型的混合智能优化算法。考虑到列车开行方案综合优化问题的复杂性,为了降低大规模客运专线网络列车开行方案求解难度,提出了基于网络划分的开行方案优化方法。更多还原

【Abstract】 Train operation plan of dedicated passenger lines is an important part of transportation organization for dedicated passenger lines. It involves rational application of the railway related fixtures and vehicle equipment. So how to make a reasonable and feasible train operation plan is a pressing and core problem of railway passenger transport. Especially, dedicated passenger lines construction has been brought into national important development strategic planning in China. The large scale construction and sustainable putting into operation of dedicated passenger lines have very important realistic meaning for accelerating urbanization and modernization of China, overall planning urban and rural areas coordination development, and meeting passenger’s travel demand. To cooperate the construction and operation of dedicated passenger lines, train operation plan making and optimization requires to be studied in time. The related theory and method should be improved.Train operation plan of dedicated passenger lines involves a lot of influencing factors and optimization processes. Among them, passenger flow forecast is the basic of making train operation plan, however, the quality of train operation plan can conversely influence the result of passenger flow forecast. This is the relationship of mutual cause and effect. So the study of integrated optimization which combines both of them is more rational. Based on the research results at home and abroad, this dissertation studies the related problems of passenger flow forecast and train operation plan. The effective solving algorithms are designed by establishing optimal models. The main research works of the dissertation are as follows:1. The commonly used passenger flow forecast methods are summarized. Because the study of induced passenger flow forecast is not systemic and mature, it is mainly studied. The improved strategy of Gravity Model is put forward. The integrated model of induced passenger flow forecast is established which combines the Gravity Model and Growth Curve Model.It can meet the requirements of induced passenger flow forecast in different periods. The induced passenger flow forecast between the main areas of Chengdu-Mianyang-Leshan dedicated passenger line is realized. 2. The passenger transfer network is the key point of the train passenger flow distribution realization. A new mixed transfer network is proposed by which can express passenger riding scheme and calculate the generalized cost conveniently. The shortest path algorithm of the network is also designed. Then based on the network, the train passenger flow distribution models are established under the condition of users equilibrium or elastic demand or capacity restraint. And the algorithms of the models are designed too.3. The problems which are closely related train operation plan are analyzed, including transport operation mode for dedicated passenger lines, economic distance of cross-line train and the operation revenue. The adaptive transport operation mode for high speed lines, the calculation method of economic distance of cross-line train and the main parts of railway operation cost and operation income are made clear. The contents, boundary conditions, principle and thought of train operation plan optimization are illustrated. Based on the train stop schedule mode, the optimization model of train stop schedule plan is established which takes the lowest travel generalized cost and passenger turnover as its optimization objective. And passenger’s economic interests and social benefit are considered in this model too. Then on the basic of the model, an optimization model of train operation plan is proposed by introducing two objective functions:operation revenue maximization and generalized impedance of train path minimization. Meanwhile, constraint conditions are added. The model is analyzed also.4. The research of intelligent optimization algorithms is summarized. And a diversity maintaining harmony search algorithm(DMHS) is proposed, which improved from two aspects. The one is the method of parameters dynamical adjustment, and the other one is updating strategy of harmony memory. Then TSP was solved by the algorithm to verify its effectiveness. After that, an optimization algorithm of train stop schedule plan is designed combined with the characteristics of the principles. Then the mixed intelligent optimization algorithm of train operation plan is designed which let DMHS embed into simulated annealing algorithm. Finally, the example result of Xian-Chengdu dedicated passenger line show that the proposed algorithm is effective.5. The integrated optimization of train operation plan based on passenger flow dynamic adjustment is studied. Based on the analysis of passenger flow forecast models, the integrated optimization model is established which takes the operation revenue maximization and generalized impedance of train path minimization as its optimization objectives. And the mixed intelligent optimization algorithm for the model is designed too. Considering the highly complexity of the problem, to reduce the difficulty of solution finding of large scale problem, the optimization method based on network partition is put forward.更多还原