Bearing Fault Detection Method Based on GA-Xgboost Machine Learning Algorithm

Bearing fault detection plays a vital role in the industry. However, current detection systems cannot identify bearing faults robustly and easily. Therefore, this paper proposes a novel algorithm to achieve fault detection of bearing based on GA-xgboost learning. SMOTE algorithm is used to deal with unbalanced data sets. Xgboost is utilized to predict bearing faults according to bearing features. Furthermore, a genetic algorithm (GA) is applied to make optimal hyperparameters of the machine learning model. The experiment results demonstrate that the algorithm can predict and detect the fault of bearing effectively with an accuracy of 96%.