Aiming at the difficulty of fault diagnosis due to the difficulty of extracting complex fault features from the vibration signals of rolling bearings,a novel method for complex fault diagnosis of rolling bearings based on optimal variable mode decomposition (VMD) and maximum correlation kurtosis deconvolution (MCKD) combined with fast spectral kurtosis algorithm was proposed.The parameters of VMD and MCKD were optimized using the improved sparrow search algorithm (ISSA),and the compound fault signal was decomposed using the optimized VMD and the effective intrinsic mode functions (IMF) were screened according to the kurtotic criterion for signal reconstruction.The optimized MCKD was used to deconvolve the reconstructed signal and enhance the fault feature.The deconvolved signal was analyzed by envelope spectrum to extract the fault feature frequency.The fast spectral kurtosis algorithm was used to process the deconvolution signal without fault characteristic frequency extracted to obtain the frequency band parameters with the most abundant fault information and perform band-pass filtering.Finally,the envelope spectrum of the filtered signal was analyzed,and the fault characteristic frequency was extracted to realize the fault diagnosis.The simulation and experimental results show that the proposed method can be used to effectively separate the complex fault and extract the fault characteristic frequency,so the complex fault diagnosis is realized effectively.