灰色补偿BP神经网络预测农机总动力—以吉林省为例

Gray Compensation BP Neural Network Prediction of the Total Power of Agricultural Machinery—Taking Jilin Province as an Example

农机总动力的预测研究对于农业机械的"供给侧"改革有着重要意义和研究价值,科学合理的预测结果对于职能部门的规划制定有着重要的指导意义。农机总动力数据具有时间序列性质,本研究应用灰色GM(1,1)模型对其进行有效的预测分析。为了提高预测的准确性,应用BP神经网络对灰色残差数据进行处理,补偿灰色预测结果,建立了相应的预测模型。实验表明:该模型对于吉林省农机总动力的预测科学有效,并对吉林省未来5年的农机总动力进行了预测,为相关政策制定提供了科学依据。

The prediction of the total power of agricultural machinery is of great significance and research value to the＂ supply side＂ of agricultural machinery. Scientific and reasonable forecast resuks have important guiding significance for the planning and development of the functional departments. The dynamic data of agricultural machinery has time se- ries properties, and the grey GM （ 1,1 ） model is used to analyze the dynamic data effectively. In order to improve the accuracy of prediction, BP neural network is used to deal with the grey residual data, and the grey prediction results are compensated, and the corresponding prediction model is established. Through experiments, it shows that the model is scientific and effective for the prediction of the total power of agricultural machinery in Jilin province. And Jilin province in the next five years, the total power of agricultural machinery to predict, to provide a scientific basis for the relevant policy formulation.