LAUSR

Gearbox compound fault pattern recognition is challenging because of its complexity and non-stationarity of the vibration signal. In this study is proposed a novel hybrid method based on narrow band interference canceller (NIC), multifractal detrended fluctuation analysis (MFDFA) and support vector machine optimized by whale optimization algorithm (WOASVM) for compound fault pattern recognition of gearbox. Specifically, the raw signal is processed by NIC to filter the deterministic signal which interferes with the fault signal, and then the multifractal features are extracted from the residual signal via MFDFA. Finally, the compound fault pattern is identified via WOASVM. Compound fault experiments of a gearbox under fixed condition and variable condition were done to evaluate the performance of the proposed method. The results show that the proposed method can effectively identify the compound faults and it outperforms other methods mentioned in this paper.