客票数据“受限”条件下旅客选择偏好估计

发布时间：2019-02-20 09:38

【摘要】：深入研究客运市场需求,是有效开展客运产品营销工作的前提.区别于既有的以问卷调查为基础的旅客选择行为研究,本文利用客票销售数据研究铁路客运产品选择偏好的估计方法.该方法考虑了部分时段部分列车在供不应求情况下,客票存根数据难以反映旅客真实选择行为的特点,根据预售期内各产品的实时余票数据,对预售期进行划分,建立了分时段的极大似然函数.针对所构建的似然函数,利用EM算法处理流失客流不可获知的问题,从而估计旅客选择偏好.通过实例,估计了徐州东至南京南方向旅客对京沪高铁客运产品的选择偏好.该方法以日常销售数据为基础,能够及时获取旅客选择偏好,具有较强的时效性,可为铁路日常营销工作提供参考.
[Abstract]:It is the premise of carrying out the marketing work of passenger transport products to study deeply the demand of passenger transport market. Different from the existing research on passenger choice behavior based on questionnaire, this paper uses ticket sales data to study the estimation method of railway passenger product choice preference. This method takes account of the fact that the ticket stub data can not reflect the behavior of passengers' real choice under the condition that part of the train is in short supply. According to the real time remaining ticket data of each product in the presale period, the pre-sale period is divided. The maximum likelihood function is established. In view of the constructed likelihood function, the EM algorithm is used to deal with the unknowable problem of the lost passenger flow, so as to estimate the passenger preference. This paper estimates the preference of Xuzhou east to south Nanjing for the choice of Beijing-Shanghai high speed rail passenger transport products. Based on the daily sales data, this method can obtain the preference of passengers in time, and has strong timeliness, which can provide a reference for the daily marketing of railway.
【作者单位】：北京交通大学轨道交通控制与安全国家重点实验室;
【基金】：中央高校基本科研业务费专项资金(2014YJS080) 中国铁路总公司科技研究开发计划课题(2013X007-3,2014X006-A)
【分类号】：U293

【参考文献】