基于元胞自动机的城市轨道交通定价双层规划  

Bi-Level Programming for Urban Rail Transit Ticket Pricing Based on Cellular Automata

作　　者：李雪岩[1] 李雪梅[1] 李学伟[1] LI Xue-yan LI Xue-mei LI Xue-wei(School ofEconomics and Management, Beijing JiaotongUniversity, Beijing 100044, China)

机构地区：[1]北京交通大学经济管理学院,北京100044

出　　处：《管理工程学报》2017年第1期155-161,共7页Journal of Industrial Engineering and Engineering Management

基　　金：国家自然科学基金资助项目(71273023);高等学校博士学科点专项科研基金资助项目(20130009110020);中央高校基本科研业务费专项资金资助项目(2014YJS059)

摘　　要：针对城市轨道交通建立了基于"通票"及"计程票"两种定价方式的双层规划模型;上层规划运用遗传算法对利润函数进行寻优;下层规划运用广义出行费用函数刻画乘客出行方式选择行为,并嵌入元胞遗传算法演化规则进行寻优。模拟结果显示:(1)乘客间的互动学习强度、对价格的敏感度是影响两种定价方式获利情况、对乘客分流情况的重要因素;(2)当乘客对票价的敏感度较低且学习进化行为较弱时,宜采用"通票"的定价方式;当乘客对票价的敏感度较高且学习进化行为较强时,宜采用"计程票"的定价方式。Striking a balance between passengers＇ satisfaction with travel demand and a certain level of profit is one of the central issues on the operation planning and pricing strategy of urban rail transit. In order to resolve this problem, the strategy of rail transit ticket pricing needs to be optimized and the passenger travel demand should be managed in a timely manner. The research of this field has received increased attention. Based on traditional bi-level based programming for rail transit, some improvements are proposed： （1） Group decision and complexity of passengers＇ mode choices are introduced; and （2） Travellers＇ behavior is described in detail and precisely. The main contents are as follows： Firstly, bi-level programming for urban rail transit pricing is established based on ＂pass ticket＂ and ＂distance-based ticket＂ in which the profit maximization oriented pricing modes follow a pricing game with sequence in upper level programming. The objective of lower level programming is the minimum of passengers＇ generalized travel cost. Secondly, passengers＇ travel mode choice model based on cellular automata is established aiming at the lower level programming. Moreover, the idea of cellular genetic algorithm is introduced to describe passengers＇ travel mode selection and the learning and evolution behaviors by using the generalized travel cost as objective function. Thirdly, the model solving process is deduced from the perspective of dynamic evolution. Afterwards, computer simulation of the model is executed to understand the influence of passengers＇ learning and evolution on rail transit operation and management departments, as well as the influence of passengers＇ ticket price sensitivity on rail transit operation and management departments and numbers of passengers in each station. In summary, there are two major findings of computer simulations conducted in this study. First, Interactive learning intension between passengers and passengers＇ sensitivity to ticket pric

关 键 词：城市轨道交通 票价 双层规划 元胞遗传算法 

分 类 号：U231.92[交通运输工程—道路与铁道工程]