Find in Library

To solve the problem that fault diagnosis accuracy of complex equipment bearings is not high due to the complexity of its structure and the environment, a cooperative algorithm for fault diagnosis of complex equipment bearings based on ensemble empirical mode decomposition (EEMD) and support vector machine (SVM) is proposed. First, the vibration signal of the bearings is decomposed by EEMD. Second, the correlation coefficient and kurtosis value are selected as the evaluation indexes for the intrinsic mode function (IMF) components after decomposition, and the weights of the parameters are set dynamically by the mean-guided weight method. Then, the IMF components are filtered by an improved genetic algorithm to obtain the optimal IMF component combination, which can effectively eliminate redundant components and retain as much fault information as possible. Next, using the orthogonality of IMF components, the energy distribution of the selected IMF components is calculated as the Eigenvector. Finally, using the advantage of accurate classification of SVM in small samples, the fault status of complex equipment bearings can be identified. The effectiveness of the algorithm model is proven by example simulation data, and the model has certain scalability and applicability in engineering.