高速铁路票价动态优化方法的研究

发布时间：2018-02-28 18:50

本文关键词： 高速铁路票价优化动态定价差异化定价旅客选择行为双层规划动态规划　出处：《北京交通大学》2017年博士论文　论文类型：学位论文

【摘要】：在基础设施以及运能配置方式不变的前提下,客票价格的优化是进一步提高运力资源利用率、增加高速铁路收益的关键。论文从当前我国铁路客票体系下的统一基准票价率和固定票价的局限性出发,对高速铁路票价的优化方法展开研究。针对不同的实际情况,分别对单列车的动态票价问题、多列车的差异化票价问题和多列车的差异化动态票价问题进行建模和求解,论文的主要研究工作如下:1、通过对高速铁路票价动态优化所解决的实际问题的描述,明确问题的定义;在理论层面将票价优化问题抽象为供需匹配问题,铁路部门提供的运输服务作为供给方,旅客作为需求方,票价则是供需匹配关系中的关键因素,进一步对供给的属性、需求的特性以及供需匹配关系进行抽象和分析,从而明确论文研究的边界;分别对单列车的动态票价问题、多列车的差异化票价问题和多列车的差异化动态票价问题建立形式化的数学模型,根据模型的结构特征梳理出三者的逻辑关系,形成问题的理论框架。2、对于起讫点间不同列车的服务属性差别不大的情况,研究单列车的动态票价问题,对列车各个运行区段分别给出调价时刻和各时刻的最优票价。首先,通过与航空以及其他国家铁路的对比,对我国铁路旅客运输的特性进行分析,作为后续建模的基础;其次,通过对客票数据的统计分析得到预售期内旅客购票时间的规律,并提出根据旅客购票规律确定调价时刻的方法;然后,在确定型需求的假设下,建立单列车的动态票价优化模型,采用拉格朗日乘子法求解;最后,通过实例验证了方法的可行性和有效性。3、在固定票价的前提下,研究起讫点间不同列车的差异化定价问题。首先,对于一定票价下的需求分布问题进行研究,在对旅客选择行为分析的基础上,提出弹性需求下的客流分配方法;其次,建立多列车差异化定价问题的双层规划模型,上层规划为以票价为自变量、以铁路部门收益最大为目标的规划模型,下层规划为给定票价下的客流分配模型;然后,改进灵敏度分析方法计算客流分配结果对于票价变化的导数,并设计基于灵敏度分析的启发式算法求解双层规划模型,经实例验证,在论文提出的差异化票价方案下客运量和铁路部门的收益均得以提高。4、对于起讫点间的多列车差异显著且需求具有明显的时变特性的一般情况,研究多列车的差异化、动态票价问题,给出不同列车在不同时间的最优票价。首先,对于该复杂的决策过程建立动态规划模型,并对模型的求解复杂度进行分析;其次,建立最优指标函数的求解模型,根据模型的结构特性推导出最优指标函数估计值的计算方法;然后,在系统状态转换过程中引入决策后状态变量,并设计基于决策后状态变量的近似动态规划算法;最后,模型和算法的可行性经实例验证,实验结果表明该票价优化方法可吸引更多旅客,并有助于提高高速铁路的收益。
[Abstract]:Unchanged in the infrastructure and capacity allocation, optimization of the ticket price is to further improve the utilization rate of resources and capacity, increase revenue. The key of high speed railway from the current railway ticket system in China under the unified standard fare rate and fixed fare limitation out of the optimization method for high speed railway fares. According to the actual situation of different, two single train dynamic pricing problem, the difference between the train fare and train the dynamic pricing problem of modeling and solving, the main works of this paper are as follows: 1, through the actual problems of high-speed railway fares dynamic optimization of the description of the defined problem at the theoretical level; will fare optimization problem into the problem of matching supply and demand, the railway departments to provide transportation services as the supply side, as the passenger demand, the fare is for The need to match the key factors in the relationship, further to the supply of property, demand characteristics and supply and demand matching relationship between abstraction and analysis, so as to clear the research boundary; two single train dynamic pricing problem, the mathematical model of differential difference train fares and multiple train fare dynamic problems the establishment of formal according to the structural characteristics of the model, sort out the logic of the relationship between the three, the theoretical framework of.2 formation, the service attributes had little differences between the start and end of the train, the fare of a single train dynamic problem, the optimal price for each train running section are given the price adjustment time and the time. First of all, with the compared to other airlines and national railway, the characteristics of railway passenger transport in our country is analyzed, as a basis for subsequent modeling; secondly, based on the statistical data of the ticket By analysis of passenger ticket pre-sale period of time, and put forward the method of determining the price adjustment time according to passenger ticket rules; then, in determining the type of demand under the assumption of establishing dynamic pricing optimization model of a single train, using the Lagrange multiplier method for solving; finally, through the example proved the feasibility and effectiveness of the.3. In the premise of fixed fare, the difference between the start and end of different train pricing problems. First, study the demand distribution problem for certain fares, based on the analysis of passenger choice behavior, the passenger flow distribution method under elastic demand; secondly, bi level planning model is established the pricing problem of multi train differences the upper, planning to fare as independent variables, programming model with the objective of maximizing the revenue of railway department, planning to lower the passenger flow assignment model under the fixed ticket price; then, improved sensitivity The calculation results for the changing price of passenger flow distribution derivative method, and design a heuristic algorithm for Solving Bilevel Programming Model Based on sensitivity analysis, it is proved that the differences in the proposed scheme of fare passenger traffic and the railway sector benefits are improved.4, has generally obvious time-varying characteristics of the train to the difference between the start and end of a significant and demand difference research train, dynamic pricing problem, given different train in the optimal price at different time. First of all, for the complex decision process to establish the dynamic programming model, and solving the model complexity analysis; secondly, to solve the model to establish the optimal index function, according to the the structure characteristics of the model push optimal index function is derived to estimate the value calculation method; then, introduces the decision in system transition process after the state variable, and based on the design decision Finally, the feasibility of the model and algorithm is verified by an example. The experimental results show that the fare optimization method can attract more passengers, and help to improve the profit of high-speed railway.

【学位授予单位】：北京交通大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：F532.5

【参考文献】