A Data-Driven Residual Life Prediction Method For Rolling Bearings

Rolling bearing is a key component of rotating machinery, which affects the reliability of the equipment. In order to make the equipment intelligent, facilitate the maintenance of the equipment, and improve the operation efficiency of the equipment, a data-driven residual life prediction method of rolling bearing is proposed, which is used to intelligently predict the remaining service life of the equipment, providing great convenience for the maintenance of the equipment. First, wavelet packet de-noising method is used to remove the noise in the original signal, and then extract the time-domain and frequency-domain features of the de-noised signal, after that PCA is used to reduce the dimension of the feature information and fuse it into the comprehensive life characteristics of the bearing. Last, with the input of the subsequent life prediction features, the residual life prediction model based on PCA-SVR is finally constructed. The experimental results show that the residual life prediction model based on PCA-SVR has low error rate and high accuracy which can accurately predict the residual life of rolling bearings.