Abstract: Aiming at the problem of damage identification of the wind turbine pitch bearings, a novel damage identification method based on optimal variational mode extraction（OVME） combined with sparse maximum harmonic-to-noise ratio deconvolution（SMHD） was proposed, aiming to extract specific signal components from composite signals. Firstly, the energy characteristic index was taken as the fitness function, and the white shark optimization algorithm was used to search for the optimal combination of influencing parameters of the variational mode extraction algorithm, so that the optimal values of the balance factor and the center frequency of the variational mode extraction were determined. Then, the variational mode extraction was used to extract the specific signal components from the vibration signals, and the extracted signal components were further deconvolved by sparse maximum harmonic-to-noise ratio to improve the signal-to-noise ratio of the signal and obtain the deconvolved signal. Finally, the envelope spectrum of the deconvolved signal was analyzed to extract the bearing damage characteristic frequency. Results show that the proposed method can accurately identify the damage characteristics of the wind turbine pitch bearings, which has a certain reference value for practical engineering.