Research On The Health Classification And Residual Life Prediction Of Rolling Bearing

Rolling bearings,as essential core components in industrial equipment,will seriously threaten life and property safety if they fail.Therefore,it is necessary to predict the remaining life of rolling bearings.In recent years,with the rapid development of sensors and computer technology,data-driven life prediction of rolling bearings has become the mainstream direction of research.This article will focus on three aspects: health status division,degradation component extraction,and residual life prediction of rolling bearings.Firstly,in response to the problem of difficulty in dividing the health status of rolling bearings throughout their entire life cycle,the root mean square value of the vibration signal is used as the degradation indicator of the bearing.The degradation indicator is adaptively segmented using time series segmentation algorithms and time series segmentation evaluation standards.The optimal segmentation method is selected as the result of dividing the health status of rolling bearings,and the segmentation point is used as the degradation threshold point for predicting the remaining life,Effectively solving the problem of difficult threshold determination under different working conditions.Secondly,based on the results of rolling bearing health state division,the degradation index series of the stationary phase and the degradation phase are selected respectively,and empirical mode decomposition is performed on them.In the degradation phase,the dynamic time warping algorithm is used to calculate the modal components or residuals with low similarity to the stationary phase,and add them together to eliminate some features that are not related to or redundant with the degradation trend,and then the bearing degradation feature components are obtained.The method proposed in this article can effectively suppress noise,obtain a single degradation trend sequence,and improve the accuracy of prediction.Finally,this article uses a deep neural network N-BEATS with residual network to predict the remaining life of rolling bearings.Considering the problem of limited historical data and high difficulty in predicting the lifespan of rolling bearings,this paper adopts a multi-step prediction structure,effectively improving the prediction accuracy.The algorithm is verified by data sets,and the results show that under different conditions,the prediction results of the method proposed in this paper are significantly improved compared with the traditional prediction model.Compared with the long-term and short-term memory neural network model,the average absolute error and the relative root mean square error are respectively increased by 12.8% and 8.2%...