Tacholess Order Tracking Method Based on STFrFT for Fault Diagnosis of Wind Turbine Gearbox Under Variable Operating Conditions

Order tracking is an effective fault diagnosis method for rotating machinery under variable speed conditions. Accurate instantaneous frequency is the premise of this method. In recent years, the tacholess method has been widely used to estimate instantaneous frequency. However, the method’s accuracy requires further improvement. Thus, this paper proposes an instantaneous frequency estimation method based on short-time fractional Fourier transform for tacholess order tracking. First, the principle and main error sources of the order tracking are analyzed. On this basis, we present three quantitative evaluation indicators of instantaneous frequency accuracy that are applicable to different scenarios. Second, an efficient and accurate instantaneous frequency estimation method based on short-time fractional Fourier transform that benefits from the first-order approximation of the linear frequency modulation kernel function to the instantaneous frequency and a series of improvement measures to avoid redundant rotation angle calculation is proposed. Finally, order spectrum is employed after tacholess order tracking to achieve fault diagnosis. The effectiveness of the proposed method is proven by the wind turbine gearbox data collected in the field. Verification results show that the proposed method is superior to the comparison methods in terms of instantaneous frequency accuracy and efficiency.