Research on the Bearing Weak Fault Signal Diagnosis Method Based on IGWO-MCKD-ROMP

In response to the issue of excessive sampling data and poor fault feature extraction performance in the fault diagnosis process of gearbox rolling bearings, a regularized orthogonal matching pursuit (ROMP) algorithm was proposed based on the maximum correlation kurtosis deconvolution (MCKD) for the diagnosis of bearing weak fault signals. Firstly, to achieve adaptive selection of parameters in the MCKD and ROMP algorithms, an improved grey wolf optimization (IGWO) algorithm was introduced. Then, the IGWO was used to denoise the original signal through MCKD processing. Finally, IGWO-ROMP was utilized to reconstruct the signal and diagnose bearing faults through envelope analysis of the signal.Verified by simulation signals and experimental results, this method can effectively extract bearing weak fault components, providing a new approach for the diagnosis of bearing weak fault signals.