Robust model predictive control of converter-based microgrids

Abstract

Converter-based microgrids are modern decentralized energy systems that integrate distributed energy resources, communication networks, and control systems. They manage energy generation, distribution, and utilization to enhance resilience and operational efficiency in modern power networks. Fourth industrial revolution technologies, such as automation, Internet-of-Things, artificial intelligence, and energy storage are integrated to create resilient, efficient, and sustainable energy systems. Model predictive control (MPC) is a promising technique for optimizing microgrid operations by considering system constraints and forecasting disturbances. However, standard MPC implementations struggle with robustness against internal and external uncertainties such as model uncertainties, measurement noise, and cyber threats, making them less effective for real-world microgrid applications. This article provides a comprehensive review of robust MPC techniques designed to enhance the reliability and security of cyber-physical microgrids. The study explores various robust MPC methodologies, including adaptive MPC, observer-based MPC, tube-based MPC, stochastic MPC, and data-driven approaches. The application of these techniques to mitigate uncertainties across different hierarchical levels of microgrid control, in particular, converter and system levels. In addition, this article examines cyber-resilient MPC approaches to mitigate cyber-attacks, including false data injection and denial-of-service attacks. Finally, emerging trends, such as AI-enhanced MPC, digital twin-based testing, and event-driven control, are outlined to support the development of next-generation robust MPC strategies for power converter-based microgrids.