华中科技大学学报（自然科学版）2011，Vol.39Issue(z2)：430-432,3.
小波包与神经网络在滚珠丝杠故障诊断中的应用
The ball screw's fault diagnosis by the wavelet packet and neural network
摘要
Abstract
A pretreatment method was proposed for ball screw's noise signal, and based on wavelet packet algorithm without frequency band derangement, we extract noise signal's character vector, by which a three layer of BP neural network was trained. Finally, the fault could be recognized in realtime by the trained weight value. The experimental results show that the ball screw s noise signals in bad lubrication and good lubrication can be distinguished correctly with the above method.关键词
故障诊断/滚珠丝杠/小波包分解/神经网络Key words
fault diagnosis/ ball screw/ wavelet packets/ neural network分类
矿业与冶金引用本文复制引用
张磊,赵晓光,张爱瑜,张承进..小波包与神经网络在滚珠丝杠故障诊断中的应用[J].华中科技大学学报（自然科学版）,2011,39(z2):430-432,3.基金项目
国家数控重大专项资助项目(2009ZX04009-013-03). （2009ZX04009-013-03）
