Fault Diagnosis of Rolling Bearings based on GA-SVM Model

The fault diagnosis of rolling bearing based on vibration signal has been conducted for many years, and various methods have emerged. The research on vibration signal analysis method usually focuses on two parts: feature selection and fault recognition. In this paper, after extracting multiple feature parameters from time domain, frequency domain, and time-frequency domain, the feature selection criterion is adopted to select significant features, which are used as the input of the support vector machine (SVM) to diagnose the faults of bearings. At the same time the parameters of the SVM are optimized via genetic algorithm (GA) to improve the classification accuracy. The proposed approach is validated by the experiment results that successfully identify three patterns of rolling bearings.