| Good failure prediction and health management play a vital role in industrial production and other areas.Accurate prediction of remaining useful life can ensure good working conditions and operational safety of machines.Existing methods for predicting the remaining useful life of rolling bearings suffer from large errors and low accuracy.In response to these problems,this paper takes the vibration signal of rolling bearings as the main research object and conducts an in-depth study on the hybrid prediction method of rolling bearing remaining useful life based on PCA-GRU,the main work is as follows:(1)In order to improve the prediction accuracy of the remaining useful life of rolling bearings with time,a model for predicting the remaining useful life of rolling bearings based on principal component analysis and stacked gated cycle units is proposed.The model has a certain improvement in prediction accuracy compared to methods such as Recurrent Neural Network(RNN),which in turn proves that the model method has good accuracy and effectiveness.(2)Aiming at the problem of high prediction errors caused by inaccurate determination of failure points and untimely classification of work stages in industry,the K-means clustering method is used to determine the FOT accurately and divide the optimal work stage,while PCA reduce the impact of redundant information related to the Multi-dimensional feature set on life prediction.The proposed model approach was tested using full-life experiments to detect the degradation time points more efficiently and the prediction accuracy was better than other methods.(3)Combining multiple model fusion strategy and ensemble learning strategy,a rolling bearing remaining useful life prediction model based on one-dimensional convolutional neural network and stacked gating cycle unit is proposed.The model improves prediction accuracy while effectively reducing the computing time of the model and ensuring the timeliness of the model. |
PDF Full Text Request