An AI-Driven Intelligent Energy Management Framework for Renewable-Integrated Smart Grids with Electric Vehicle and Power Electronics Coordination
Abstract
Energy-management strategies are needed to be smart, intelligent, and adaptive to coordinate the variable renewable-energy generation with the fluctuating electric vehicle-charging demand and grid operating conditions in renewable-integrated smart grids. In this work, an AI based intelligent energy-management framework for renewable integrated smart-grid with a particular emphasis on the coordination of electric vehicles and prediction of operational efficiency with renewable energy resources is developed. A secondary smart-grid dataset was analysed that includes information of electric vehicle count, charging demand, arrival and departure time, renewable energy generation, grid load, peak-demand status, renewable utilization ratio, load-balance index, real-time data exchange, and operational-efficiency data. Major energy-demand, renewable-generation, and grid-operation patterns were characterized by descriptive statistical methods. Correlation and comparative analyses were performed to investigate the relationships between electric vehicle charging demand, renewable-energy use, grid load, peak-demand situations, and load-balancing performance. Machine-learning classification models: Logistic Regression, Decision Tree, Random Forest and XGBoost were added to predict operational-efficiency categories using the available smart-grid variables. To evaluate the model performance, accuracy, precision, recall, F1-score and confusion matrices were used, and for feature-importance analysis, the most influential factors related with smart-grid operational efficiency were identified. The suggested framework offers a well-defined way to incorporate EV charging behaviour, renewable energy availability, and grid operating conditions in a single data-driven energy-management framework. The results are intended to help coordinate EV charging more efficiently, utilizing renewables more effectively, controlling peak demand, balancing load and making smart decisions in future smart-grid settings