摘要
为预测云南省农机总动力的发展变化趋势,提出一种将GA算法、LM算法与BP神经网络相结合的农机总动力预测方法,克服了BP神经网络易陷于局部极小的缺点。选取1985-2015年云南省农机总动力数据作为样本,建立GA-LM-BP神经网络模型进行仿真预测,结果表明:该模型的平均相对误差为0.313 362%,明显优于BP神经网络的0.926 674%、LM-BP神经网络模型的0.654 053%和GA-BP神经网络模型的0.493 122%,具有较好的预测精度。在此基础上,对云南省2016-2020年农机总动力的发展趋势进行了预测,结果表明:2 0 1 6年农机总动力达3 4 3 9.4 9万k W,超过云南省农业厅预测的3 4 0 9万k W,2 0 2 0年云南省农机总动力达3 952.78万k W,为云南省农机化的发展规划提供了理论依据。
In order to predict the development trend of the total power of Yunnan Province,s agricultural machinery,this paper puts forward a method of combining GA algorithm,LM algorithm and BP neural network to predict the total power of agricultural machinery,which overcomes the shortcoming that BP neural network is easy to fall into local minimum.Selection of 1985-2015 in Yunnan Province,the total power of agricultural machinery as the sample data,GA-LM-BP neural network model is established for simulation and prediction. The results show that the average relative error of the model is 0. 313 362%,significantly better than the BP neural network model's 0. 926 674%,LM-BP neural network model's 0. 654 053% and GA-BP neural network model's 0. 493 122%,has good prediction accuracy. On this basis,the development trend of the total power of agricultural machinery in Yunnan province 2016-2020 years are predicted. The prediction results show that the total power of agricultural machinery in 2016 to reach 34 million 394 thousand and 900 kilowatts,exceeding the Yunnan Provincial Department of agriculture forecast of 34 million 90 thousand kilowatts,the total power of agricultural machinery in Yunnan Province in 2020 amounted to 39 million 527 thousand and800 kilowatts,providing a theoretical basis for the development of Agricultural Mechanization in Yunnan province.
出处
《农机化研究》
北大核心
2018年第4期47-52,共6页
Journal of Agricultural Mechanization Research
基金
重庆市重点产业共性关键技术创新专项(cstc2015zdcy-ztzx80003)
关键词
农机总动力
GA-LM-BP模型
预测
total power of agricultural machinery
GA-LM-BP model
prediction