Research On Prediction Methods For The Remaining Life Of Gearbox Based On Condition Monitoring

Wind power is considered to be the most competitive power generation technology at the price and is leading the global energy transformation.As wind turbines installed capacity increases,the operation and maintenance workload of wind farm is also increasing,and the gearbox,as the key component of wind turbine and the device with the highest failure rate in wind turbine,is usually installed at the top of the wind turbine.Maintenance is difficult,and high operation and maintenance costs seriously affects the economic benefits of wind farm.Therefore,monitoring the operating state of the gearbox and predicting accurately its remaining life provide an important reference and basis for the predictive maintenance of wind turbines.In this thesis,a real-time remaining life prediction method based on the theory of the wiener process and kernel density estimation is studied to predict the remaining life of two key components(gears and bearings)in wind turbine gearbox.On this basis,considering the effects of gear and bearing degradation rate,a performance degradation model of two components considering stochastic dependence is established,and the validity of the model is verified by experiments.The main elements are as follows:(1)A real-time remaining life prediction method based on wiener process and kernel density estimation theory is proposed to solve the problem that traditional models need to assume parameter distribution and lead to large prediction error.The nonparametric kernel density estimation method is used to estimate the distribution of the degradation rate of the gear and bearing parts of the gear box system respectively.The remaining life probability density function of the gear and bearing is obtained from the perspective of probability statistics,and the remaining life of the gear and bearing is dynamically predicted by realtime updating of the degradation rate.(2)Considering the stochastic dependence of gears and bearings in the degradation process,based on the wiener performance degradation model,the correlation relationship model of gear and bearing degradation process is established by using the degradation rate correlation method.It is verified by experiments that this method can model the degradation process of gear box and obtain more accurate remaining life prediction results.