Role of artificial intelligence in next-generation renewable power systems

The use of renewable energy (RE) is increasing to meet carbon neutrality goals. The wide application of RE requires specific energy production forecasting, control, and operation; power dispatch optimization to reduce expenses while addressing operational constraints; efficient system regulation to ensure a balanced power supply; and electricity markets with auctioning and trading decision-making related to RE. The fluctuations in RE generation make the power system difficult to operate. This discontinuous behavior of wind power makes it challenging to maintain supply and demand for electrical power in actual operating conditions; however, conventional power generation sources may be necessary to meet demand, increasing electricity costs. This Review highlights the capability of artificial intelligence (AI) based techniques to support the operation of a renewable power system. We analyze the capabilities of machine learning, reinforcement learning, and deep learning techniques for enabling power system predictions, dispatch, control, and market operations to support the adoption of RE. We also highlight the feasibility of these methods for various functional issues. Finally, we analyze the potential for trends in RE power system methods and advancements to address related operational difficulties, including the increasing decentralization of RE implementations, variations in energy storage mechanisms, and increasing market complexities.