Abstract: As the development of new-generation power systems continues to progress, power electronic equipment has become an increasingly vital component in ensuring the stable and efficient operation of these systems. However, with design requirements characterized by large-scale complexity, diversity, and high-performance standards, conventional design approaches based on expert knowledge, as well as existing electronic design automation（EDA） tools, are increasingly insufficient to meet practical demands. The rapid advancement of artificial intelligence（AI） technologies, particularly in the domains of large language models, reinforcement learning, and graph neural networks, has introduced novel and effective methodologies for addressing the challenges inherent in power electronics circuit design. A comprehensive review of recent research developments in the application of AI to power electronics design and control is presented, with a focus on core areas such as chip layout optimization, circuit topology enhancement, control strategy formulation, parameter optimization, and intelligent design assistance. Through a comparative analysis of the technical features, data requirements, and performance advantages of various AI-based approaches, the major challenges currently facing the field are identified and discussed in this study. These include the limited availability of high-quality training data, insufficient generalization ability of AI models, and challenges related to the seamless integration of such models into practical engineering applications. The current state of research is summarized and potential directions for future investigation are proposed to conclude the paper. The aim is to provide a solid theoretical foundation and practical guidance for advancing the intelligent design of power electronic systems through the effective integration of artificial intelligence techniques.