Aiming at the problem that state degradation model structure hypothesis in the process of remaining useful life prediction of the wind turbine gearbox, a real-time remaining useful life prediction method combining kernel density estimation and stochastic filtering theory was proposed. This method used the kernel density estimation method from the data itself to estimate the probability density function of the gearbox continuous degradation state, and obtained the degraded state probability density function based on real-time state monitoring data, and then used the real-time condition monitoring data to update the stochastic filter recurrence model parameters to predict the real-time residual life of the gearbox. Effectiveness of the proposed method was verified by the test of gearbox. © 2020, Editorial Department of CIMS. All right reserved.
