Optimizing the Analysis of Energy Plants and High-Power Applications Utilizing the Energy Guard Ensemble Selector (EGES)

Abstract

Accurate performance assessment of energy plants and high-power electrical systems is challenging due to the dynamic nature of parameters like energy output, voltage levels, and load factors. This study introduces the Energy Guard Ensemble Selector (EGES), a machine learning-based algorithm designed to enhance predictive accuracy and reliability in power electronics. EGES employs a dynamic model selection approach, leveraging classifiers such as Random Forest, Support Vector Machine, Gradient Boosting Machine, K-Nearest Neighbors, and Logistic Regression. By using KNN to evaluate real-time electrical conditions, EGES dynamically selects the most suitable model to predict key metrics such as energy output (MW), efficiency (%), fault rates, and transformer capacity (MVA). Experimental results show that EGES outperforms individual models with an accuracy of 93.5%, precision of 91.5%, recall of 92.7%, and an F1-score of 92.1%, demonstrating its robustness in handling fluctuations in electrical parameters. EGES proves to be a reliable tool for improving predictive accuracy and functional dependability in high-power electrical systems.