Abstract: Faced with the severe challenges of energy shortage and ecological environment deterioration, renewable energy has played an increasingly important role. Wind power generation, as one of the three major new energy sources, has been widely promoted and applied in the world in recent years. As a key component in capturing wind energy and converting it into mechanical energy, the blades of wind turbines are of great importance. Once the blade is damaged, it will seriously affect the service life and power generation efficiency of wind turbines. Therefore, the defect detection technology of fan blade has become the focus and mainstream of current research. The defect types of fan blades are divided into layer failure, composite layer failure, bonding failure and component failure according to their damage degree. On the basis of summarizing the basic principle and research progress of non-destructive defect testing (WBT), this paper makes an in-depth analysis and comparison of these non-destructive defect testing technologies, which promotes the understanding of the integrity of wind turbine blade detection and provides strong support for the sustainable and healthy development of wind power industry.

Key words: blade, wind turbine, defect detection, deep learning, ultrasonic detection

摘要： 在能源短缺和生态环境恶化的背景下，可再生能源日益重要，风力发电作为三大新能源之一，近年来得到了全球广泛应用，而风力发电机叶片作为关键部件，承担着捕获风能并转化为机械能的核心功能，其损坏将显著降低发电效率和缩短设备使用寿命，因此，风机叶片的缺陷检测技术已成为研究的热点，主要针对图层失效、复合层失效、胶接失效及组件失效等多种损伤类型。总结了多种无损缺陷检测技术（WBT）的基本原理、研究进展以及各自的优缺点，从适用场景、技术优势与局限性等方面对不同检测方法进行了系统归纳与分析，结合风力发电工程的实际需求，提出了各类检测技术的应用建议，并展望了风机叶片缺陷检测领域的未来发展方向，旨在为风力发电行业的持续健康发展提供有力支持。

关键词: 叶片, 风力发电机, 缺陷检测, 深度学习, 超声波检测