Feature extraction based on generalized permutation entropy for condition monitoring of rotating machinery

Defective rotating machinery usually exhibits complex dynamic behavior. Therefore, feature representation of vibration signals is always critical for condition monitoring machinery. Permutation entropy (PeEn), an adaptive symbolic description, can measure complexities signals. However, PeEn, which compresses all the information into a single parameter, may lack capability to fully describe dynamics Afterward, multiscale PeEn (MPeEn) put forward coping with nonstationarity, outliers and artifacts emerging in In MPeEn, set parameters serves different time scales. Nonetheless, average procedure MPeEn withhold local destroy internal structures To overcome deficiencies this paper proposes generalized (GPeEn) by introducing orders lags PeEn. GPeEn, signal converted matrix rather than parameter. Moreover, minimal, maximal values serve briefly conditions Next, numerical experiment proves that proposed method performs better skewness, kurtosis, characterizing Lorenz model. Subsequently, compared investigating gear roll-bearing containing types severity faults. The results show outperforms other four methods distinguishing between faults