Fault prediction of rolling bearings based on PCA and PSO-SVM

In order to predict the failure of rolling bearing, a prediction model based on PCA and PSO-SVM is proposed. First, the main component is extracted from the experimental data of rolling bearing of Case Western Reserve University. Then the optimal parameters of SVM model are calculated through the PSO algorithm. Finally, the optimal parameters and data set are input into the SVM model for failure prediction. The results show that the prediction accuracy of PCA-PSO-SVM model achieves 97.2%, which is 19.4% higher than the standard SVM model, resulting in higher prediction accuracy.