摘要
目的提出一种新的BP神经网络非线性组合预测模型,对河北省2016-2023年农机总动力进行预测。方法基于一元非线性回归以及BP神经网络(BP-ANN),建立一种新的BP神经网络非线性组合预测模型,处理1978-2009年河北省农机总动力数据,建立预测模型并进行比较。结果误差分析表明,该非线性组合预测模型的拟合平均绝对误差为1.286%,低于一元非线性回归和传统BP神经网络。利用此模型对2010-2015年河北省农机总动力进行检验预测,预测结果与实际结果有很好的一致性,为农机总动力的预测提供了一条新的途径。结论此模型有较高的预测精度,并用此模型预测了河北省2016-2023年农机总动力数值,预测结果表明,在未来几年河北省农机总动力将保持高速增长,2023年将达13 610.69万kW。
Objective This paper proposes a new BP neural network nonlinear combination forecast model,and forecasts the total power of agricultural machinery of Hebei province from 2016 to 2023.Methods Based on the unitary nonliner regression mode and BP neural network（BP-ANN）,a new BPANN combined with nonliner regression mode was proposed.The total powers of agricultural machinery from 1978 to 2009 in Hebei Province were investigated.Results The results of error analysis showed that mean absolute error of proposed nonlinear combined prediction model was 1.286%,lower than those of one-variable liner regression mode and traditional BP-ANN.Values of agricultural machine power of Hebei province from 2010 to 2015 were forecasted.The prediction values based on the model were in good agreement with the original values.Conclusion This model has higher accuracy to predict the total power of agricultural machinery from 2016 to 2023 for Hebei province.The prediction results suggest that total power of agricultural machinery will maintain swift growth tendency in the near future for Hebei Province,and it will be 1 361 069 000 kW in 2023.
出处
《河北北方学院学报（自然科学版）》
2018年第1期45-49,共5页
Journal of Hebei North University:Natural Science Edition
基金
张家口市科学技术和地震局项目(1101016B)
关键词
农机总动力
一元非线性回归分析
人工神经网络
预测
total power of agricultural machinery
unitary nonlinear regression analysis
artificial neural network
prediction
