1. Designing an optimal microgrid control system using deep reinforcement learning: A systematic review
- Author
-
Noer Fadzri Perdana Dinata, Makbul Anwari Muhammad Ramli, Muhammad Irfan Jambak, Muhammad Abu Bakar Sidik, and Mohammed M. Alqahtani
- Subjects
Microgrid ,Control system ,Deep reinforcement learning ,Engineering (General). Civil engineering (General) ,TA1-2040 - Abstract
Microgrid systems play a pivotal role in the integration of renewable energy sources and enhancing electrical grid resilience. Deep Reinforcement Learning (DRL), a subset of artificial intelligence, holds the potential to revolutionize the control and management of microgrids. This systematic review aims to provide a comprehensive assessment of the current state of research on designing microgrid control systems using DRL. In this review, an overview of microgrid systems is presented, along with their components, and the inherent challenges in control and management. By systematically categorizing the chosen articles, a summary and critical evaluation of the primary discoveries, methodologies, and contributions in this area are tabulated. Additionally, an assessment of research methodology quality is conducted, while common patterns, hurdles, and emerging technologies are identified. Furthermore, significant issues and unanswered research queries related to the implementation of DRL in microgrid control systems are highlighted. Furthermore, it aims to gain insight into the pros and cons of utilising DRL algorithms such as exploration, off-policy reinforcement learning (RL), inverse RL, goal-conditioned RL, and multi-agent RL. This study will be valuable for researchers or practitioners wishing to conduct research on optimal design of renewable based microgrid systems, particularly in the application of DRL algorithms for microgrid control systems. It is also expected, that this study will be useful in identifying research gaps or the state of the art in the potential relationship between DRL and microgrid control system.
- Published
- 2024
- Full Text
- View/download PDF