摘要
机器的二次信号往往因机器故障产生大量的冲击、摩擦以及运行转速的不稳定、负荷的变化导致非平稳信号的产生.对非平稳信号分析,付氏变换效果不佳,需要研究这类信号的局部时频特征,提取瞬变信息方能准确地诊断.本文介绍处理非平稳信号的新型工具——小波分析、短时付氏变换两种时频分析方法.最后用小波分析、短时付氏变换和付氏变换对机器的实测振动信号进行分析.说明了小波分析、短时付氏变换作为时频分析方法对处理非平稳信号比付氏变换优越.
Non-stationary signal arises from mechanical shock,friction,unstable speed and changed load. Fourier transform doesn't fit its processing. So,it is necessary to study its local time-frequency feature and extract its transient information. In this paper,two kinds of new time-frequency analysis approaches for non-stationary signal processing are introduced,which are wavelet transform and short-time Fourier transform. These two approaches and Fourier transform are applied to analyse vibration signal. From this paper,conclusions are obtained to show that time-frequency analysis is superior to Fourier transform in non-stationary signal processing.
出处
《振动．测试与诊断》
EI
CSCD
1993年第4期47-55,共9页
Journal of Vibration,Measurement & Diagnosis
基金
国家自然科学基金
机械结构强度和振动国家重点实验室资助的项目
关键词
傅里叶变换
瞬变信息
机器诊断
short-time Fourier transform,wavelet analysis,transient information,non-stationary signal,machinery diagnosis
