Evolution identification method of rolling bearing compound fault based on VCESS index and bidirectional dichotomy approach SEACKgram

The time point of compound fault evolution plays an important role in bearing healthy stage division and life prediction. However, most existing fault monitoring methods do not pay attention to the identification of compound fault evolution time points. In order to solve this problem, considering the influence of random impact on time series, an evolution identification method of compound faults of rolling bearings based on evolution trend index and bidirectional dichotomy approach square envelope autocorrelation correlation kurtosis (SEACK)gram is proposed. By constructing the SEACK, the variation coefficient of envelope spectrum SEACK (VCESS) is analyzed, and the compound fault evolution trend index VCESS is put forward. According to the characteristics that the evolution time points of different types of faults have great changes corresponding to the evolution trend indicators, the evolution trend of compound faults is monitored and an approximate evolution time point is identified. Furthermore, SEACKgram method combined with bidirectional dichotomy approach strategy is used to accurately locate the evolution time point of compound faults. The effectiveness of the proposed method is verified by the open full-cycle rolling bearing life fault data experiment, which provides a more accurate basis for health state division and life prediction.