铁路旅客购票需求预测模型研究

Forecasting Model for Railway Passenger Ticketing Demand

铁路旅客购票需求是列车票额分配的重要依据,各区间旅客的购票需求即为旅客的出行需求,在预售期间有不同的趋势规律。本文基于旅客购票历史数据,分析同一季节不同区间旅客的购票分布,提出区间旅客平均购票强度的概念,利用购票强度来描述预售期各OD票额的动态需求。将购票提前天数、购票渠道、单次购票人数、出行OD和票价等旅客购票行为的关键特征变量作为影响预测日期的属性向量,提出了非线性回归支持向量机的预测模型,对预售期每日的区间旅客购票需求进行预测。最后通过算例对模型进行了可行性验证。

Railway passenger ticketing demand（RPTD） is the critical basis for a train tickets allocation, RPTD trend in each section performed differently in the pre-sale period. Based on history ticketing data, this paper put forward the concept of average section ticketing intensity（TI） by analyzing the ticketing distribution during the pre-sale period. The TI was applied to describe the dynamic ticket demand of each origin-destination（OD）. The critical characteristic variables of passenger ticketing behavior, such as advance ticketing days, purchasing channel, the number of tickets for once purchasing, travel OD and ticket price, were taken as the attribute vectors which affected forecasting date. A prediction model of nonlinear regression support vector machine was proposed for forecasting the daily tickets demand for each OD in the pre-sale period. Finally, the feasibility of the model was verified by an example.