时序混合模型预报及其在铁路货运收入中的应用

Time series combined model and its application to the income forecast of railway station

本文以某火车站记载的,以往的月货运收入序列,采用潘迪特—吴(Pandit-Wu)建议的混合模型的方法建模,采用滚动式预报,获得了较满意的结果。此外,还举例说明了原始记录真实性的重要性,并提出了值得研究的建模中的计算精度问题。

According to the monthly income of railway freight series recorded for a certain station and the modeling method of combined model suggested by Pan-dit-Wu, a time series combined model is established in this paper.The forecast of rolling calculation of this model is satisfactory. Some examples in this paper show that the precision of original records is very important. The paper also suggests that the calculation precision in the modelling is worth studying further.