农机总动力组合预测模型研究-基于密度算子

Research on Combination Forecasting Model of Agricultural Machinery Total Power Based on Density Operator

农机总动力的需求具有增长性、波动性及非线性的特征。采用基于密度算子的组合预测方法,对陕西省农机总动力进行预测。在确定单一预测模型的基础上,计算各模型的偏离度和准确度,进而确定聚类组的准确度和密度加权向量,建立基于密度算子的农机总动力组合预测模型,以及基于离差系数法和Shapley值法的组合预测模型。拟合结果表明:建立的密度算子组合模型的各项误差评价指标都低于选定的单一预测模型和基于离差系数法和Shapley值的组合预测模型,具有很好的预测效果。

The demand for total power of agricultural machinery has the characteristics of growth, fluctuation and non-linearity. The combined forecasting method based on density operator is used to forecast the total power of agricultural machinery in Shaanxi Province. On the basis of determining the every single prediction model, the deviation and accuracy of each model are calculated, and then the accuracy of clustering group and density weighting vector are determined, and the combined forecasting model of agricultural machinery total power based on density operator is established. and then combined forecasting model based on deviation coefficient method and Shapley value method is established. The fitting results show that the error evaluation indexes of the density operator combination model are lower than those of the selected single prediction model and the combination prediction model based on the deviation coefficient method and Shapley value, which has a good prediction effect.