Huang Liping, Lu Kai, Zeng Liang
School of Mechanical Engineering, Xi'an University of Science and Technology, Xi'an 710054, China.
School of Mechanical Engineering, Xi'an Jiaotong University, Xi'an 710049, China.
Sensors (Basel). 2025 Jul 17;25(14):4466. doi: 10.3390/s25144466.
The wind turbine blade is subject to multi-source impacts, such as bird strikes, lightning strikes, and hail, throughout its extended service. Accurate localization of those impact sources is a key technical link in structural health monitoring of the wind turbine blade. In this paper, a single-sensor impact source localization method is proposed. Capitalizing on deep learning frameworks, this method innovatively transforms the impact source localization problem into a classification task, thereby eliminating the need for anisotropy compensation and correction required by conventional localization algorithms. Furthermore, it leverages the inherent coding effects of the blade's material and geometric anisotropy on impact sources originating from different positions, enabling localization using only a single sensor. Experimental results show that the method has a high localization accuracy of 96.9% under single-sensor conditions, which significantly reduces the cost compared to the traditional multi-sensor array scheme. This study provides a cost-effective solution for real-time detection of wind turbine blade impact events.
风力涡轮机叶片在其长期服役过程中会受到多种来源的冲击，如鸟击、雷击和冰雹。准确确定这些冲击源的位置是风力涡轮机叶片结构健康监测中的关键技术环节。本文提出了一种单传感器冲击源定位方法。该方法利用深度学习框架，创新性地将冲击源定位问题转化为分类任务，从而无需传统定位算法所需的各向异性补偿和校正。此外，它利用叶片材料和几何各向异性对来自不同位置的冲击源的固有编码效应，仅使用单个传感器即可实现定位。实验结果表明，该方法在单传感器条件下具有96.9%的高定位精度，与传统的多传感器阵列方案相比，显著降低了成本。本研究为风力涡轮机叶片冲击事件的实时检测提供了一种经济高效的解决方案。
