A New Interpretable Learning Method for Fault Diagnosis of Rolling Bearings

摘要: In modern manufacturing processes, requirements for automatic fault diagnosis have been growing increasingly as it plays a vitally important role in the reliability and safety of industrial facilities. Rolling bearing systems represent critical part most applications. view strong environmental noise working environment rolling bearing, its vibration signals nonstationary nonlinear characteristics, those features are difficult to be extracted. this article, we proposed new intelligent method with unlabeled data by using convolutional neural network (CNN) fuzzy $C$ -means (FCM) clustering algorithm. CNN is first utilized automatically extract from signals. Then, principal component analysis (PCA) technique used reduce dimension extracted features, two components selected feature vectors. Finally, FCM algorithm introduced cluster derived space identify different types bearing. The results indicate that newly can achieve higher accuracy than other existing literature.