Fault Prediction for Rotating Bearing United Genetic Algorithm Optimize Variational Mode Decomposition and Support Vector Machine

The vibration signal of rotating bearing is continuous and non-stationary. The fault information contained in the signal can not be reflected directly due to the excessive noise background of the operating environment. Based on the above characteristic and shortcomings, a new fault prediction method is proposed in this paper. Firstly, aiming at the disadvantage of modal aliasing caused by improper parameter selection of variational mode decomposition(VMD) algorithm, a parameter optimization model based on genetic algorithm(GA) is established to search the best combination of parameters globally. Due to the small dataset of early rotating bearing failures, the support vector machine (SVM) method based on parameter optimization is used for fault prediction, and its prediction accuracy is as high as 90%, which provides a new idea for rotating bearing fault prediction.