摘要
本文提出了应用差分计算和句法模式识别方法识别机械振动和噪声谱中谱峰的新方法。此方法数学概念清楚,使用效果好。本文中对此方法做了详细的讨论。由于引入了模拟人工分析思路的自学习功能和模糊数学方法,因而能有目的地对各种谱峰加以识别。试验研究表明,此方法能提取和识别出振动噪声谱中与机械状态有关的谱峰,有利于实现机械设备的状态监测和故障诊断的自动化。
A new method for recognizing peaks in vibration and noise spectrum,which uses difference operatorsand syntactic pattern recognition,is discussed in detail in this paper.The method is simple theoreticallyand good practically.Since a self-study function to simulate human thought and a method of fuzzy mathe-matical analysis are introduced,the method can get some characteristics of artificial intelligence and detectvarious kinds of spectral peaks purposely.Practical uses show that the peaks,which are relateing with thecondition of a machine,in vibration and noise spectrum can be automatically picked up and recognized withthe method.Therefore,the use of the method is helpful to realizing the automation of diagnosis of machin-ery.
出处
《振动工程学报》
EI
CSCD
1993年第3期255-262,共8页
Journal of Vibration Engineering
关键词
机械振动
故障诊断
噪声谱
波形分析
mechanical vibration
fault diagnosis
noise spectrum
waveform analysis
pattern recognition
