Explainable and Generative AI for Smart Grids: A Unified Framework and Research Directions

Abstract

Smart grids are becoming more complicated with distributed energy resources, dynamic load changes, real-time operating requirements. The conventional black-box artificial intelligence (AI) models lack in trust, transparency and compliance with regulations, thereby limiting their use in practical deployment. To resolve these issues, this paper suggests a conceptual unified GenAI-XAI framework for smart grids. Explainable Artificial Intelligence (XAI) offers an interpretable and reliable decision-making ability for existing scalability and real-time integration related issues. Recent advances in Generative AI (GenAI) complement XAI by supporting in scenario modeling, intelligent decision assistance, and synthetic data generation. The study presents a comparative analysis of XAI methods and a systematic taxonomy of GenAi applications in anomaly detection, load forecasting, energy optimization, and predictive maintenance The proposed conceptual framework help to improves adaptability, transparency,, and operational efficiency, thereby facilitating the development of a sustainable, resilient, and smart grid systems.