摘要
在对铲式成穴器穴孔长度进行二次通用旋转组合试验设计的基础上,引入基于人工神经网络的数据处理方法。分析结果表明,BP神经网络对试验结果拟合模型的相对剩余误差比回归模型拟合的相对剩余误差小,拟合精度高于回归模型,为铲式成穴器成穴质量的试验研究提供了新的数据分析方法。
Based on general quadric rotation experimental design for punch length of spade soil opener, a data processing method based on BP neural network has been introduced to take better advantage of experimental information. The results show that the relative residual standard deviation of the fitted value of BP neural network is lower than that of the regression model. So a new analysis method is provided for the experimental study of the punch quality of the spade soil opener.
出处
《农机化研究》
北大核心
2004年第3期134-136,共3页
Journal of Agricultural Mechanization Research
基金
教育部博士点专项基金资助项目(1999003001)
关键词
神经网络
铲式成穴器
穴孔长度试验
应用
artificial inthelligence
spade soil opener
experiment analysis
punch length
BP neural network
