组合模型在农业用水量预测中的应用

Application of Combined Model in the Prediction of Agricultural Water Consumption

在分析灰色模型与人工神经网络模型特点的基础上提出了灰色神经网络组合预测模型,并以最小二乘法确定了2个单一模型权重系数的分配。结果显示,灰色神经网络组合预测模型的最大、最小及平均误差均明显优于2个单一模型,有效提高了预测的精度。

Based on analyzing the characteristics of grey model and artificial neural network model,the combined grey-neural network prediction model was put forward and the assignment of weight coefficients of 2 single models was ascertained by least square method.The results showed that the maximum,minimum and average errors of combined grey-neural network prediction model were less than that of the 2 single models and the prediction precision was enhanced effectively.