Mitigating supra harmonics in the textile industry using renewable resources: The PCA-WHOGBDT technique for improving grid stability and power quality

Abstract

Supra harmonics are high-frequency harmonics that can cause a variety of problems in power grids (PGs). This paper proposes a hybrid technique to reduce supra-harmonics in the power grid. The methodology employed in this study combines Principal Component Analysis (PCA) and Wild Horses Optimization (WHO)-based Gradient Boosted Decision Trees (GBDT). The study analyzes network frequency variations, DC voltage development, and voltage noise. It enhances LCCL and LCL filters and evaluates the strategy's effectiveness on weak and stiff grids using real-time data. The proposed PCA-WHOGBDT successfully reduces supra-harmonics to 0.2% of the fundamental in weak grids, outperforming typical levels. It mitigates high-frequency harmonics in both grid types. The PCA-WHOGBDT strategy provides an effective solution for reducing supra-harmonics in hybrid-fed grids, contributing to improved power system quality and stability.