With large-scale integration of renewable generation and distributed energy resources, electric power systems are confronted with emerging operational challenges, such as growing complexity, increasing uncertainty and aggravating volatility.
Meanwhile, more and more data is becoming available owing to the widespread deployment of smart meters, smart sensors and upgraded communication networks.
As a result, advanced artificial intelligence (AI) techniques, notably reinforcement learning (RL), have attracted surging attention for enabling data-driven control and decision-making in complex smart grid systems.
To bridge the gap between power and AI domains, Dr. Xin Chen, an assistant professor in the Department of Electrical and Computer Engineering at Texas A&M University, and his team have authored a paper about using reinforcement learning in smart grid applications.
Chen recently received recognition for this paper, “Reinforcement Learning for Selective Key Applications in Power Systems: Recent Advances and Future Challenges," from the Institute of Electrical and Electronics Engineers (IEEE) Transactions on Smart Grid journal. It ranked third in the top five outstanding papers among over a thousand articles published in 2020-2022.
“I was thrilled and honored to receive this acknowledgement as it is a very prestigious recognition from one of the leading academic journals in the field of power and energy systems,” Chen said. “Smart grids are the future of power systems. We envision the power grid to be more intelligent, resilient and sustainable. This journal is dedicated to the critical technology and applications in this field.”
Smart grids are the future of power systems. We envision the power grid to be more intelligent, resilient and sustainable.
The paper presents a comprehensive and structural overview of the RL methodology, from basic concepts and theoretical fundamentals to state-of-the-art RL techniques. It elaborates on the RL application to power system decision-making with key application examples and points out critical challenges as well as promising future directions.
“In power systems, we have various decision-making problems,” Chen said. “For example, we need to determine how to optimally operate power generators and power networks. AI-assisted control and decision-making algorithms can serve as advanced innovative tools to help solve these issues.”
Since its initial publication in IEEE Transactions on Smart Grid in 2022, this paper has garnered significant attention from both academia and industry.
“It's a rapidly emerging field,” Chen said. “Our intention in writing this paper was to provide interested engineers and researchers with a comprehensive guide, enabling them to explore the dynamic, interdisciplinary realm of using reinforcement learning in power system applications."
This work represents a collaborative effort led by Chen, involving Dr. Guannan Qu from Carnegie Mellon University, Dr. Yujie Tang from Peking University, Dr. Steven Low from the California Institute of Technology and Dr. Na Li from Harvard University.
Chen also received an invitation to present this work in an upcoming IEEE Power and Energy Society webinar.