Advanced Study on Structural Health Diagnosis and Maintenance for Floating Wind Turbines Using Computer Vision

Global offshore wind capacity has now surpassed 50 GW and is projected to reach 264 GW by 2050, highlighting the pivotal role of floating wind in the future of clean energy. Given the complexity of marine environments, intelligent diagnostics for floating turbines are crucial for improving operational efficiency, reducing costs, and ensuring robust and sustainable energy production. This paper presents a structural damage detection framework for floating wind turbines, integrating computer vision with advanced artificial intelligence technologies. First, a dataset is constructed through industry collaboration and open-source collection. Then, to optimise the YOLOv7 algorithm, SE attention mechanisms and WISE-IoU loss functions are incorporated, which significantly enhance the accuracy of surface damage detection. Experimental results indicate that the mAP (mean Average Precision) increases from 82.44% to 86.24% compared to the original YOLOv7. Finally, a deployment approach and an example are provided to use the diagnostic framework as a portable application. This enables real–time on–site analysis, enhances detection timeliness, and reduces maintenance costs. It allows for immediate issue identification and adaptation to diverse environments.