Adaptive parameter-matching method of SR algorithm for fault diagnosis of wind turbine bearing

The fault diagnosis of wind turbine bearings is challenging because of the heavy background noise and changes in wind speed. Stochastic resonance (SR) is an effective method of detecting fault signal from noise. However, the benefit of SR is seriously limited by the system parameters and the frequency of the input signal. A novel fault diagnosis method for wind turbine bearings, combining an adaptive SR algorithm that is based on quantum particle swarm optimization (QPSO) and frequency conversion based on frequency information exchange (FIE), is proposed. First, the frequency information of the fault characteristic signal is exchanged with the reference frequency by FIE, which can eliminate the limitation of the frequency band. Then, the SR system parameters are optimized by QPSO to avoid blind parameter selection. The signal after FIE is processed by the optimized SR system. The results of case study show that under the same input signal, the proposed method can achieve better signal-to-noise ratio and response amplitude than can the traditional double-side band modulation method and an SR method that is combined only with an optimization algorithm.