Rotor misalignment detecting by novel adaptive time–frequency mode decomposition and parametric stochastic resonance

Misalignment faults frequently occur in a variety of rotor faults but extracting the characteristic is a challenge especially when the equipment operating under time-varying speed conditions and with standing various noise interferences. A novel adaptive time–frequency mode decomposition method is proposed to extract the feature of the misalignment fault. The theoretical study is carried out to successfully decompose the feature modes in the presence of strong noise with time-varying speed conditions. An instant crest factor is proposed as a discriminant basis for identifying feature order. Then, the fault feature is further enhanced by parametric stochastic resonance. In addition, the effectiveness of the method is verified by experimental study. Experimental results demonstrate that the proposed method can extract the characteristics of rotor misalignment faults successfully and will be of practical value in engineering.