基于改进CenterNet算法的风机叶片损伤检测识别技术

Wind Turbine Blade Damage Detection Recognition Technology Based on Improved CenterNet Algorithm

为了对风力发电机组叶片损伤状态进行有效检测,提出一种基于CenterNet目标检测算法的风机叶片损伤检测识别技术。该技术选取DLA-60特征提取网络作为CenterNet算法的骨干网络,并在DLA-60网络中引入注意力引导数据增强机制,提升检测算法的精度。优化后风力机叶片损伤检测识别模型的检测精度为88%,较原始算法提升了2.6个百分点,且检测时间基本与原网络持平,具有较强的精确性和实用性。

In order to effectively detect the damage state of wind turbine blades, this paper proposes a wind turbine blade damage detection and recognition technology based on the CenterNet target detection algorithm. The DLA-60 feature extraction network is selected as the backbone network of the CenterNet algorithm, and the attention-guided data enhancement mechanism is introduced in the DLA-60 network to improve the accuracy of the detection algorithm.Experiments show that the optimized wind turbine blade damage detection and recognition model has a detection accuracy of 88%, which is 2.6% higher than the original algorithm, and the detection time is basically the same as the original network, which has great practicability and accuracy.