基于加权马尔科夫-ARIMA修正模型的区域物流需求预测

程元栋，喻可欣，李先洋

安徽理工大学经济与管理学院，安徽 淮南  232001

摘要：为准确预测区域物流需求，采用自回归移动平均（autoregressive integrated moving average，ARIMA）模型建立具有线性关系的时间序列，考虑时间外的非线性影响因素，基于加权马尔科夫模型修正残差状态，构建加权马尔科夫-ARIMA模型，以我国1990—2021年月度货运周转量为物流需求数据来源，验证加权马尔科夫-ARIMA模型的预测精度。结果表明：单一ARIMA模型和加权马尔科夫-ARIMA模型对12期货运周转量预测结果的平均绝对百分误差分别为3.15%、2.22%，后者的预测精度优于前者。

关键词：ARIMA模型；加权马尔科夫模型；物流需求；预测

The regional logistical demand forecast based on weighted Markov-ARIMA modified model

CHENG Yuandong, YU Kexin, LI Xianyang

School of Economics and Management, Anhui University of Science and Technology，Huainan 232001, China

Abstract:To accurately predict regional business logistical demand, an auto-regressive integrated moving average(ARIMA) model with a linear relationship is established for series of timing, in the meantime, non-linear influences outside of timing are also considered, then, the residual statuses are modified based on the weighted Markov model, finally the weighted Markov-ARIMA model is constructed. To test the predicted accuracy of the weighted Markov-ARIMA model, China′s monthly trucking turnovers from the year 1990 to 2021 as the source of business logistical data are employed. The results show that the average absolute percentage errors of the single ARIMA model and the weighted Markov-ARIMA model for the 12-period trucking turnovers forecasting results are 3.15% and 2.22% respectively, and the forecasting accuracy of the modified model is better than that of the single ARIMA model.

Keywords:ARIMA model; weighted Markov model; logistical demand; forecast