The application of morphology analysis and RFFSVM to intelligent fault diagnosis on the bearing of ships

Support Vector Machine SVM is widely applied to fault diagnosis of machines. However, this classification method has some weaknesses. For example, it can not separate fuzzy information, particularly sensitive to the interference and the isolated points of the training samples. In view of the problems mentioned above, a random forest fuzzy SVM multi-classification algorithm (RFFSVM) has been put forward. This paper focuses on the study of the application of the Morphology Analysis and the theory RFFSVM (MA-RFFSVM) to fault diagnosis on the bearing of ships. Simulation experiments show that the algorithm has better anti-interference ability and classification effects than others. Consideration should be taken into account that it can be further applicable to the diagnosis on other mechanical faults of ships.