摘要
将模糊技术与RBF神经网络相结合并应用于机床热误差建模中,构建了基于模糊RBF神经网络的数控机床热误差模型;以某龙门导轨磨床主轴箱系统为实例,将模糊RBF神经网络建模方法运用到主轴箱系统热误差建模当中。通过与BP神经网络建模方法进行对比,验证了模糊RBF神经网络建模方法无论是在建模效率、建模鲁棒性还是模型的补偿效果方面均优于传统的BP神经网络建模方法,该方法对提高数控机床加工精度具有重要的意义。
A thermal error model of NC machine tool was established by combining fuzzy technology with RBF neural network.Illustrated by an example of gantry rail grinder spindle box system,the fuzzy RBF neural network modeling approach was applied to thermal error modeling of spindle box.Fuzzy RBF neural network modeling approach is proved much better than traditional BP neural network modeling approach in modeling efficiency,modeling robustness and compensation effect.This approach has important significance for improving machining accuracy of NC machine tools.
出处
《机械设计与研究》
CSCD
北大核心
2013年第5期71-74,80,共5页
Machine Design And Research
基金
国家自然科学基金资助项目(51175161)
关键词
模糊技术
径向基函数
模糊RBF神经网络
热误差建模
fuzzy technology
radial basis function
fuzzy RBF neural network
thermal error modeling
