Computer Vision based Operational Condition Monitoring and Risk Detection of Offshore Wind Farm Converter Valve

This paper focuses on the digital intelligent operation and maintenance technology in the scene of offshore wind power converter stations and proposes a condition monitoring and risk detection algorithm based on computer vision. Due to the environmental particularity of offshore converter stations and the important position of converter stations in DC transmission, it has important practical significance and research value for the multi-aspects monitoring and inspection of converter stations. In recent years, with the development of deep learning, many detection algorithms based on computer vision have been implemented in industrial scenes, replacing manual inspection work in complex and dangerous conditions, and improving equipment safety and inspection efficiency. Based on the improved yolov9 model and two-stage visual detection method proposed in this paper, the detection and identification of a variety of faults and risks such as personnel safety operation, smoke and fire, equipment dirt and corrosion are recognized in multiple scenarios of converter station, which provides technical support for the intelligent operation and maintenance of digital converter station of offshore wind power platform, helps to reduce the operation and maintenance cost and improve the operation efficiency of converter station.