Compound fault diagnosis method for rolling bearings based on the improved symplectic period mode decomposition

The symplectic period mode decomposition (SPMD) method can accurately extract the periodic pulse components in a signal, which is an effective method for the single fault diagnosis of rolling bearings. However, in the case of composite faults in rolling bearings, especially under strong background noise, the periodic pulse signals are often weak, which makes it difficult to extract the pulse components with different periods, thus limiting its application in the diagnosis of composite faults. An improved symplectic period mode decomposition (ISPMD) method was proposed to deal with this regard. The method firstly adopts the combination of the strengthen operate subtract operate enhancement technique and minimum noise amplitude deconvolution method to reduce the noise in the signal and enhance the period pulse to accurately estimate the fault period. Then, the periodic segment matrix was constructed and the symplectic geometry period component was obtained by the symplectic geometry similarity transformation and the periodic impact intensity. Finally, the residual signal was decomposed by iteration and the symplectic geometry period components with different periods were obtained. The experimental results show that ISPMD can accurately extract the periodic impulse components, which is an effective method for composite fault diagnosis of rolling bearings. © 2024 Chinese Vibration Engineering Society. All rights reserved.