基于灰色BP神经网络的农业机械总动力预测

为预测宁夏地区农业机械化水平的发展变化趋势,提出一种将灰色预测模型与BP神经网络有效结合的农业机械总动力预测方法。在BP神经网络的数据预处理阶段融入灰色预测理论,建立基于灰色BP神经网络的农机总动力预测模型,并选取1991-2014年宁夏回族自治区农业机械总动力数据作为样本,利用该模型进行仿真预测,结果表明:该模型具有较高的预测精度,其平均相对误差仅为0.18%,明显优于灰色GM(1,1)模型的3.5 0%和标准BP神经网络的0.2 9%。

Prediction of the Total Power of Agricultural Machinery Based on Grey BP Neural Network

To predict the development trends of agriculture mechanization in Ningxia province, the method combined grey prediction model and BP neural network is proposed.By incorporating grey prediction theory in data preprocessing stage of BP neural network can construct the prediction model of the total power of agricultural machinery based on grey BP neural network.Besides, we choose the data of total power of agricultural machinery in Ningxia province from 1991 to 2014 as a sample, and using the model to predict the simulation.The result of simulation show that this model has high prediction accuracy, which average relative error is up to 0.18%, better than the grey GM(1,1) model of 3.50% as well as the BP neural network of 0 .29%.