Remaining useful life prediction of wind power gearbox based on KECA and Wiener process
The gearbox is a key equipment of wind turbines,and once its performance deteriorates to a failure state,it can cause serious safety hazards.In order to accurately predict the remaining useful life of wind power gearbox,a remaining useful life prediction method based on kernel entropy component analysis and Wiener process is proposed.In data preprocessing,random forest algorithm is used to eliminate outliers and abnormal data,and Pearson algorithm is used to select multiple features with high correlation with gearbox degradation.By using the kernel entropy component analysis method to perform principal component analysis in high-dimensional space,the principal components with a large amount of information retention are selected to achieve the goal of data dimensionality reduction.Then,the residual life of the wind power gearbox is predicted using the Wiener process.Taking the actual data of a wind farm in Hebei province as an example,the results show that when using 3000,4000,and 5000 points for prediction,the prediction errors of the proposed method are 12.72%,10.52%,and 6.05%,respectively,which is significantly better than the comparison method.
wind power gearboxremaining useful life predictionkernel entropy component analysisWiener processrandom forest algorithm
万方数据
维普
