关键词: 冲击脉冲法, 滚动轴承, 神经网络, 自修正系数

Abstract: Shock pulse method (SPM) is introduced. This method is applied to rolling bearing condition monitoring and fault diagnosis. Neural networks based parameter self-modification strategy using for SPM are presented. The system has been working in industry site on-line. The long period of perfect running has proved that the system can highly improve the accuracy of fault diagnosis.

Key words: Neural networks, Parameter self-modification, Rolling bearing, Shock pulse method