物流需求预测的效率和精度直接影响物流系统的有效规划、管理,考虑到大多区域经济变量具有非线性与波动性的特征,传统的线性建模和预测技术难以适应物流需求预测的精度要求.因此,提出基于SVR残差修正自回归的区域物流需求预测模型.首先采用残差自回归模型拟合序列的线性部分;然后对拟合后的残差序列采用SVR方法再次提取其非线性信息,从而提高模型的预测精度;最后以广东省公路货运量的预测为例说明模型的有效性.为满足社会物流需求,有效地进行物流规划、管理,提供了良好的决策支持.%In this paper,we proposed a regional logistics demand forecasting model based on SVR-revised residual auto-regression model.First we fitted the linear part of the sequence using the residual auto-regression model,then again extracted the nonlinear information of the residual sequence after the fitting using the SVR method in a bid to increase the forecasting aecuracy of the model,and at the end,taking the roadway freight volume of Guangdong for instance,demonstrated the effectiveness of the model.
展开▼
