Detection of Bearing Fault in Induction Motor Using Multi-parameter Optimized Resonance Sparse Signal Decomposition

When the induction motor bearing fails, the stator current signal not only has weak fault features related to fault information, but also includes plenty of strong background noise, which increases the difficulty of fault detection. In order to effectively extract the fault features, this paper proposes a multi-parameter optimized resonance sparse signal decomposition (RSSD). The proposed method does not depend on frequency range, but divides the spectrum through resonance, and overcomes the limitation of traditional RSSD methods that rely on manual experience to set important parameters such as quality factors. The novelty of this method lies in the introduction of gorilla troops optimizer (GTO) algorithm to automatically select quality factor Q, weight factor A and Lagrange multiplier \(\mu \) . Firstly, with the minimum fitness function as the goal, GTO is used to optimize the selected parameters. Secondly, RSSD is used to obtain the best resonance component and power spectral density (PSD) carried out to extract the bearing fault feature frequency. The experimental results show that the proposed method is more effective in detecting bearing faults than the traditional method.