Application of an optimized minimum noise amplitude deconvolution method in bearing fault diagnosis

In the bearing fault diagnosis, detecting the cyclic pulse component in the vibration signal is the key to fault feature extraction. However, the complex operating environment often makes the weak pulse component masked by background noise and vibration interference. Therefore, a bearing fault diagnosis method based on parametric adaptive minimum noise amplitude deconvolution (PAMNAD) is proposed. This method uses the improved grey wolf optimization (IGWO) algorithm to optimize the minimum noise amplitude deconvolution (MNAD), and determines the optimal values of the filter length, the number of iterations and the noise ratio. The optimized MNAD shows higher adaptability and robustness under different working conditions. The experimental results show that this method has remarkable effectiveness and superiority in actual bearing fault signal processing.