基于统计特征的轨道交通站点乘降量预测算法研究

针对现有预测方法因未考虑城市轨道交通站点乘客的随机性,致使乘降量预测精度不高的问题,提出一种基于统计特征的客流量预测方法。依据日期、时段、天气、突发事件等因素将历史数据进行分类。建立基于乘降量统计特征的分布模型,根据其预测客流的统计特征,结合随机数产生算法,产生的随机数即为客流乘降量预测值,最后结合算例予以说明,证明该模型的可行性。

Prediction Algorithm of the Passengers on and off at Tram Stations Based on Statistical Characteristics

This paper proposes a statistical characteristics-based ridership-predicting method to solve the low predication accuracy of the number of passengers on and off as the current prediction method does not consider passenger randomness of urban rail stations. The method covers the following steps: divide and classify historical data according to such factors as dates, time intervals, weathers and emergencies; create a distribution model based on the statistical characteristics of the number of passengers on and off; and calculate random numbers in accordance with the statistical characteristics of the predicted passenger flow and the random generation algorithm. The random number is the predicted number of passengers on and off, which along with the example demonstrates that the distribution model is workable.