Artificial intelligence–assisted data-driven control of power electronics systems

Data-driven solutions enabled by artificial intelligence (AI) have shown great potential in power electronics. Compared with conventional control methods, data-driven control methods are flexible and efficient and have fewer requirements regarding system physical knowledge. As for control-related applications, this chapter presents AI-based control approaches to power electronics applications, covering control fundamentals, data-driven principles, practical examples, and outlooks. It starts by discussing the control fundamentals from a data-driven perspective. Then, relevant AI tools, including metaheuristic methods, fuzzy logic, and machine learning methods, are presented. Specific examples are analyzed in control optimization, a fuzzy logic–based controller, a neural network-based controller, and a reinforcement learning controller. Finally, future outlooks on the state of the art in this synergistic topic are summarized.