Resonance Sparse Decomposition Method for Optimizing Quality Factor of Frog Leaping Algorithm

Abstract

Due to the complex and varied application scenarios of rolling bearings in engineering practice， early rolling bearing failures are easily masked by high-intensity background noise， and traditional fault feature extraction methods are not ideal. Aiming at the rigorous manual selection of the quality factors of the existing resonance sparse decomposition method， a frog leaping algorithm optimization parameter model is proposed， and the quality factors are optimized by the frog leaping algorithm， then the optimal quality factor combination is input into the resonant sparse decomposition model for processing， the low resonance component is obtained， and the low resonance component is subjected to Hilbert demodulation analysis. The maximum kurtosis value of the low resonance component is set as the goal， and iterative optimization of frog leaping is performed to finally identify the fault frequency.

Keywords

• Rolling bearing
• Resonance sparse decomposition
• Quality factor
• Frog leaping algorithm