At present, the feature extraction of frequency signal based on empirical mode decomposition (EMD) has been widely studied and applied in fault diagnosis of rolling bearings. However, there are still some shortcomings in fault diagnosis based on EMD. Therefore, a fault diagnosis method based on the combination of EMD and target feature selection (TFS) is proposed in this paper. The method firstly analyzes the fault signal through EMD and extracts the fault features. Then, it removes the redundant features and optimizes the feature subsets by using TFS. TFS selects the most effective feature for each target sample space through filtering evaluation criteria and heuristic search strategy, thereby effectively improving the accuracy and efficiency of fault diagnosis.

In order to make up for the deficiency of traditional single diagnosis in rolling element bearing fault diagnosis application, eliminate a large amount of redundant information and improve the classification effect of the aliasing mode, based on comprehensive analysis of the respective advantages of fuzzy set and tree search, this paper presents a joint rolling bearing fault diagnosis method based on tree-inspired feature selection and FS-DFV (Fuzzy Set and Dependent Feature Vector). The dependent feature vectors (DFV) can dig deeper the essential differences of the faults and improve the fault accuracy. By establishing the heuristic tree model, the tree type heuristic feature search strategy is designed, and the excellent feature selection criteria based on the density clustering with noise are proposed, and the conventional feature selection model is improved. In addition, fuzzy set are used to process the problem of extracting aliasing patterns in the DFV, and fuzzy membership is used to guide subsequent feature extraction of the alias modes. The proposed method is compared with the other four fault diagnosis methods. The experimental results show that the proposed method can effectively improve the diagnostic efficiency of the rolling bearing.