CACMD And MCKD Methods For Rolling Bearing Fault Diagnosis In Wind Turbines

Wind turbines as a wind power conversion of important equipment,primarily wind energy-based new power network is effectively accelerating the transformation of China’s energy structure,promoting the development of "net zero emissions".As the main support component of wind turbines,rolling bearings have a serious impact on the stable operation of the machine if there is an internal failure.For this reason,there is a very serious need to study bearing failure problems for reducing the safety risks of wind turbines and ensuring the operational efficiency of the turbines.With the rolling bearings of wind turbines as the research object,this paper is centred on the signal processing properties of the ACMD and MCKD algorithms for feature extraction research.Further details of the study are as follows.(1)To solve the problem that the current Adaptive Chirp Mode Decomposition(ACMD)lacks the ability to extract periodic pulses and over decomposes modal components,a correlated Gini index(CG)is introduced by using the Gini index(GI)characteristics,and then a CACMD method integrating CG with ACMD Mode Decomposition is proposed.Compared with VMD,TVF-EMD,FEEMD and original ACMD,CACMD has more advantages in decomposition accuracy and calculation speed.(2)A new weighted comprehensive fault feature index(EKE)is proposed to achieve the optimal selection of the key parameters L and M of the MCKD filter,and then the MCKD parameters are adaptively optimized based on the improved Multi Verse Optimizer(IMVO)algorithm.Compared with PSO and MVO iteration results,the accuracy and reliability of IMVO-MCKD method are verified.(3)The IMF component evaluation criterion is designed for selecting CACMD components,and the CACMD-MCKD rolling bearing fault diagnosis method based on the principle of cragginess and correlation coefficient is proposed and applied to experimental data of real single faults as well as compound faults.The results show that the method can adequately eliminate unimportant information in non-smooth signals,thus enhancing the extraction of fault features.On the whole,according to the non-stationary,strong shock and other characteristics of signal oscillation occurring during the operation of wind turbines,and the rolling bearings in gearboxes are studied.Work of this paper provides certain ideas for the identification and diagnosis of bearing faults.