Fault Diagnosis Method of Planetary Gearbox Based on ISSA-VME and Multi-scale Fuzzy Entropy

Abstract

Abstract To solve the problem of noise interference and difficulty in feature extraction of vibration signals of planetary gearbox, propose a fault diagnosis method of planetary gearbox based on improved sparrow search algorithm optimized variational mode extraction (ISSA-VME) and multi-scale fuzzy entropy (MFE). The vibration signal was decomposed by ISSA-VME, and the multi-scale fuzzy entropy of the original signal was calculated according to the IMF component. The fault feature set composed of MFE was applied to the SVM classifier optimized by the improved sparrow search algorithm for fault pattern recognition. The experimental results demonstrate that compared with ICEEMDAN and VMD methods, ISSA-VME has a better effect and higher efficiency on vibration signal decomposition. The identification rate of missing teeth, cracks, wear, and other faults of planetary gearbox sun gear by the proposed method reaches 97.50%.