Fault diagnosis of rolling element bearing based on principal component analysis of acoustic signal

Fault diagnosis of using acoustic signal generated by a machine has a lot of advantages,such as easier signal collection,non-contact measurement,no requirement of sticking sensors in advance,no in-fluence of the collection system on the machine,easier actualization of early forecasting and on-line mo-nitoring,and able to be widely used in the occasion where it is difficult to collect vibration signal.How-ever,it is more difficult to extract the characteristic signal due to the effect of environment noise.Ther-efore,the principal component analysis is employed to preprocess the original acoustic signal.Then,En-velope analysis based on Morlet wavelet transform and spectrum analysis are applied to extract the fault characteristic vector.The proposed method has been applied to the fault diagnosis of rolling element be-aring.The experimental results show that the proposed method in this paper is effective to diagnose early fault of rolling element bearing.