PQD Classification Method based on Iterative Filtering Kaiser Window S-Transform and NRBO-XGBOOST

Résumé fourni par la source

To address the inadequate time-frequency analysis of the S-transform and the low accuracy of traditional Power Quality Disturbance (PQD) classification methods, this paper presents a classification method based on Iterative Filtering Kaiser Window S-Transform (IFKST) and Newton-Raphson-Based Optimizer to optimize Extreme Gradient Boosting (NRBO-XGBoost). Firstly, the iterative loop filtering interval positioning algorithm extracts frequency feature points of disturbance signal. Then the K-S transform is applied to each feature point. Finally, 17 time-frequency features are extracted from the IFKST resulting matrix, and the hyperparameters of Extreme Gradient Boosting are optimized using the Newton-Raphson-based optimizer. The 17 disturbance features are then input into the model to achieve PQDs classification. Experimental results show that this method can classify 21 types of power quality disturbances with high accuracy in noisy environments.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.