Emerging Trends in Electrical Engineering: Integrating Smart Grid Technologies, Automation, and Artificial Intelligence for Sustainable Power Systems

Abstract

Smart grids are increasingly important for sustainable power systems because they combine digital communication, automation, renewable energy integration, and intelligent control. However, maintaining grid stability remains challenging due to decentralized producer-consumer interactions, renewable energy variability, and dynamic demand behavior. This study developed an artificial intelligence-based framework for classifying smart grid operating states as stable or unstable using reaction time, power balance, and price elasticity features. The Smart Grid Stability Augmented Dataset was used, containing 60,000 observations and 14 variables. Twelve input features were selected, while the categorical variable stabf was used as the classification target. The continuous stability score stab was excluded to prevent data leakage. Data preprocessing included missing-value checking, duplicate inspection, label encoding, feature scaling, and stratified 80:20 train-test splitting. Several machine learning and deep learning models were evaluated, including Logistic Regression, Decision Tree, Random Forest, Support Vector Machine, Gradient Boosting, XGBoost, and Artificial Neural Network. The results showed that nonlinear and ensemble-based models outperformed baseline classifiers. Tuned XGBoost achieved the best performance, with accuracy of 0.9848, precision of 0.9847, recall of 0.9916, F1-score of 0.9882, and ROC-AUC of 0.9990. Overall, the findings demonstrate that AI-driven stability prediction can support automated monitoring, early instability detection, and sustainable smart grid decision-making.