Abstract: To enhance the accuracy of transformer fault diagnosis based on machine learning, a transformer fault diagnosis method using NRBO-XGBoost based on Dissolved Gas Analysis (DGA) is proposed. The Extreme Gradient Boosting (XGBoost) model is selected. The Newton-Raphson-based optimizer (NRBO) is combined with XGBoost to iteratively search for the optimal model parameters. In each iteration, the performance of the current solution is evaluated, and the XGBoost model is retrained using the optimal parameters obtained. A significant improvement in model performance is observed by comparing the results before and after optimization. The performance of the NRBO-XGBoost method is evaluated through a case study, demonstrating the effectiveness of the proposed method for transformer fault diagnosis, with good convergence and high accuracy.

Key words: transformer, fault diagnosis, Newton-Raphson-based optimizer, Extreme Gradient Boosting