Bearing Fault Diagnosis based on Independent Component Analysis and Optimized Support Vector Machine

This study concerns with fault diagnosis in rolling bearings using discrete wavelet transform (DWT), statistical parameters, independent component analysis (ICA) and support vector machine (SVM). The features for classification are extracted through using statistical parameters combined with energy obtained through the application of Db2-discrete wavelet transform at the fifth level of decomposition. After feature extraction, ICA is employed to select the relevant features. Finally an optimized SVM based on particle swarm optimization (PSO) is used for bearing fault decision. The obtained results proved the effectiveness of the proposed methodology for bearing faults diagnosis.