A review of intelligent control strategies for energy management systems in microgrids

Microgrids (MGs) are increasingly pivotal for integrating distributed energy resources (DERs) such as photovoltaic and wind systems to enhance sustainability, resilience, and efficiency. However, the variability of renewables necessitates advanced Energy Management Systems (EMS) to ensure reliable, cost-effective, and stable operations. This review evaluates EMS frameworks namely, centralized, decentralized, and distributed, and control strategies including Model Predictive Control (MPC), Artificial Intelligence (AI)-based methods, stochastic and robust models, Multi-Agent Systems (MAS), classical optimization, bio-inspired heuristics, and hybrid algorithms. Findings reveal that centralized EMS achieves global optimization but is limited by scalability and single-point vulnerabilities, decentralized EMS enhances autonomy but sacrifices holistic coordination, and distributed EMS, particularly MAS, offers scalability and resilience when supported by robust communication. While MPC and AI-driven models provide strong adaptability and forecasting accuracy, they are computationally demanding, whereas stochastic and robust methods improve reliability under uncertainty but face efficiency challenges. Hybrid EMS emerges as the most effective, integrating complementary approaches to optimize costs, renewable utilization, emissions reduction, and resilience. The study also underscores the growing vulnerability of intelligent EMS to cyberattacks, emphasizing the need for cyber-resilient architectures with detection and self-healing mechanisms. Furthermore, emerging communication innovations, including IoT networks and 5G technologies, are highlighted as critical enablers of scalability, reliability, and secure coordination in future EMS. Advancing adaptive AI-driven hybrid EMS, resilient communication infrastructures, and experimental validation will be key to achieving intelligent, secure, and sustainable microgrid operations.