The global trend indicates that overall wind energy production, both onshore and offshore, will increase drastically in the next decade. Therefore, presently, much effort is focused on optimizing the operation and maintenance of wind turbines, since these are quite challenging and cost‐intensive. To aid or even completely fulfill a specific inspection task, an automated solution is proposed in this paper. The prototype is built on an M300 drone platform from DJI Technology Co. and is presented here. It requires a single, additional 2D‐LiDAR sensor mounted on an upwards frame. The proposed control and path planning algorithms have been tested in the AirSim simulation environment, as well as in local model airfields and at real onshore and offshore wind turbines. As a result, a comprehensive sequential‐phased mission is presented, which reduces the total time required for the inspection routine to approximately 14 min, representing about half the time an expert pilot may need for the same task. Additionally, a platform prototype that may be deployed on a ship's deck for a safe landing is presented. It guarantees instant adhesion upon contact and avoids unwanted drone backlash due to sudden and unexpected ship movement during the landing approach. Further work will focus mainly on additional offshore flight probes, optimizing the landing platform, and tuning the flight algorithms.

中文翻译：

---

基于激光雷达的风力涡轮机转子叶片自动无人机检查

全球趋势表明，陆上和海上风能总产量将在未来十年大幅增加。因此，目前，很多努力都集中在优化风力涡轮机的运行和维护上，因为这些工作非常具有挑战性且成本密集。为了帮助甚至完全完成特定的检查任务，本文提出了一种自动化解决方案。该原型机基于 DJI Technology Co. 的 M300 无人机平台构建，如下所示。它需要安装在向上框架上的单个附加 2D-LiDAR 传感器。所提出的控制和路径规划算法已在 AirSim 仿真环境、本地模型机场以及真实的陆上和海上风力涡轮机中进行了测试。因此，提出了全面的分阶段任务，将检查例行程序所需的总时间减少到大约 14 分钟，大约是专家飞行员完成相同任务所需时间的一半。此外，还提出了一个可以部署在船舶甲板上以安全着陆的平台原型。它保证接触时立即粘附，并避免由于着陆过程中突然和意外的船舶运动而导致无人机不必要的反弹。进一步的工作将主要集中在额外的海上飞行探测器、优化着陆平台和调整飞行算法。