Fault diagnosis of rolling bearing based on EEMD and CSSVM
                                                                   LIANG Zhihua1, CAO Jiangtao1, JI Xiaofei2
(1.School of Information and Control Engineering, Liaoning Shihua University, Fushun 113001, China;2.School of Automation, Shenyang Aerospace University, Shenyang 110136, China)

Abstract: Aiming at the problem that in the optimization of support vector machines, datadriven rolling bearing fault diagnosis, most of them are classified by support vector machine, however, the traditional support vector machine classification method is easy to fall into local optimum, and it is impossible to carry out fault diagnosis accurately, the feature selection of rolling bearing vibration signal and the optimization method of support vector machine were studied. The disadvantages of the support vector machine optimization by using genetic algorithm and particle swarm optimization were pointed out. In order to improve the accuracy of rolling bearing fault diagnosis, a cuckoo search algorithm based on Levy flight was introduced to find the optimal parameters of support vector machine. First, ensemble empirical mode decomposition was used to process signal data, and then the root mean square of the intrinsic mode function was put into the support vector machine optimized by cuckoo search algorithm to train and test this model. The results indicate that the proposed method can be used to analyze and diagnose the measured signals, and the location and severity of the faults can be accurately identified. The superiority of the algorithm is verified by comparison with traditional optimization methods.

Key words: ensemble empirical mode decomposition(EEMD); cuckoo search(CS); support vector machine(SVM); fault diagnosis