Unknown fault diagnosis of rolling bearing based on improved GWO-LightGBM

In view of misjudgment of unknown new faults of rolling bearing affects bearing safety and maintenance efficienc，a fault diagnosis model based on improved gray wolf optimization （GWO） and light gradient boosting machine （LightGBM） was proposed to realize high precision discrimination about the known and unknown faults.The time domain，frequency domain and wavelet domain features were extracted separately from the vibration signal of the rolling bearing to avoid the lack of feature extraction at a single scale.The GWO-LightGBM model with unknown new fault diagnosis mechanism was designed，and the improved gray wolf algorithm with Halton sequence and simulated annealing strategy was constructed to realize the effective optimization of model parameters.The experimental results showed that the average recognition rate of the model for known and unknown faults was 99.57%.The average recognition rates for 10 times random experiments were 21.98%，17.00% and 9.27% higher than logistic regression （LR），K-nearest neighbor （KNN） and support vector machine （SVM），respectively.The comparative experiments verified the effectiveness and superiority of the model，which can identify known or unknown new faults with high accuracy.