To address the problem of inaccurate selection of demodulation frequency band and difficulty in fault diagnosis that may arise when using the unbiased autocorrelation kurtosis index,the square envelope negentropy is introduced into Autogram,and an improved Autogram method using square envelope negentropy index is proposed to select the optimal demodulation frequency band.The square envelope negentropy is less affected by noise and can measure the impact of period signal.Firstly,the bearing vibration signal is decomposed into different frequency bands by maximum overlapping discrete wavelet packet transform.Then the square envelope negentropy of each frequency band is calculated,and the optimal frequency band which has the maximum square envelope negentropy value is selected.Finally,the bearing fault type is diagnosed by analyzing the square envelope spectrum of the optimal demodulation frequency band.The analysis results of fault bearing experimental signals by improved Autogram,Autogram,and fast kurtogram method showed that the improved Autogram method is more reasonable in the selecting demodulation frequency band,and more obvious in extracting fault frequency,and more accurately in diagnosing bearing faults.
关键词
平方包络负熵/改进Autogram/最优解调频带/故障诊断
Key words
square envelope negentropy/improved Autogram/optimal demodulation frequency band/fault diagnosis
维普
万方数据