Fault feature extraction of wind turbine rolling bearing based on MOBWO-MCKD
Aiming at the problem that the vibration signal of wind turbine bearings affected by strong background noise and other equipment excitation sources leaded to the difficulty of feature extraction for early bearing weak fault features,a wind turbine bearing fault diagnosis method based on maximum correlation kurtosis deconvolution(MCKD),optimized by multi-objective beluga whale optimization(MOBWO)algorithm was proposed.Firstly,based on the powerful global and local search capabilities of MOBWO,the key parameters of MCKD were optimized,and the optimal parameter combination of MOBWO was obtained.Secondly,the optimized MCKD was employed to process the original signal by deconvolution operation for eliminating the influence of background noise and other equipment excitation sources and highlighting the bearing periodic pulse signal.Then,the envelope spectrum method was used to process the deconvolution signal to perform the extraction of bearing fault characteristic frequency,and the obtained fault characteristic frequency values were compared with the theoretical calculation values for diagnosis.Finally,in order to validate the effectiveness of MOBWO-MCKD,the experiments were conducted on the actual collected inner and outer ring fault data of wind turbine bearings.The results show that the fault feature extraction method based on MOBWO-MCKD can effectively eliminate the background noise and other excitation source interference of the early bearing weak fault features.The inner ring signal envelope spectrum shows the inner ring failure frequency fIR=125.87 Hz and 2fIR=251.74 Hz.The envelope spectrum of the outer ring signal can be seen as the outer ring failure frequency that fOR=84.47 Hz,2fOR=168.94 Hz,3fOR=253.41 Hz,which has a certain application value for the extraction of early weak fault characteristics of wind turbine bearings in practical engineering.
万方数据
维普
