Enhanced Rolling Bearing Fault Diagnosis Combining Novel Fluctuation Entropy Guided-VMD with Neighborhood Statistical Model

Variational Mode Decomposition (VMD) provides a robust and feasible scheme for the analysis of mechanical non-stationary signals based on variational principle, but this method still has no adaptability, which greatly limits application in bearing fault diagnosis. To solve problem effectively, paper proposes novel fluctuation entropy (FE) guided-VMD essential characteristics impulse signals. The FE reported not only considers order amplitude values also variation amplitude, hence it can comprehensively characterize transient rolling signal. On basis establishing FE, FE-based fitness functions are then conducted, after mode number balance parameter be adaptively determined. Meanwhile, an adaptive neighborhood statistical model is developed to further reduce noise component containing information so as highlight periodic more significantly improve diagnostic accuracy. Simulation case show that research effective quite accurate separation feature enhancement. Compared with traditional VMD current common diagnosis methods, proposed obvious advantages comprehensive utilization enhanced