基于偏好顺序集的非补偿式铁路旅客选择行为估计方法

精细化需求预测是铁路提高收益的重要基础.本文考虑了铁路产品属性对于旅客选择行为的影响,构建具有一定偏好顺序的产品集合来表征不同类型的旅客.在此基础上,利用客票存根及余票数据,建立预售期内旅客选择过程的极大似然函数,求解得到不同旅客类型的出现概率.首先通过仿真算例验证了模型的计算可行性,然后通过实证数据,估算了北京南至上海虹桥方向不同旅客类型在各时段的出现概率,进一步统计得到旅客在不同时段下铁路旅客buy-up行为概率及属性敏感度的变化,并提出一些可用于优化售票策略的建议.

A Method Considering Non-compensation Principle to Estimate Railway Passenger Choice Behavior Based on the Preference Choice Set

The refinement of demand forecast is an important basis for the railway to increase revenue. This paper considers the influence of railway product attributes on passenger choice behavior. We use the preference queue of products to describe a particular types of passenger. The maximum likelihood function of passenger selection process in the pre-sale period is formulated based on sale data to calculate the arrival probability of different types of passenger. In this paper, the effectiveness of the model is verified by the simulation example. Furthermore, the arrival probability of the different types of passenger from Beijing South Railway Station to Shanghai Hongqiao Station is estimated by empirical data. Further statistics is utilized to get the buy-up behavior and the sensitivity to attributes in different periods, and the result can be used to optimize the ticket selling strategy.