基于改进Otsu_Canny算法的光伏组件故障定位研究

Research on the Fault Location of PV Modules Based on an Improved Otsu_Canny Algorithm

为了提高光伏系统在线故障诊断的效率,提出一种基于改进遗传算法的Otsu_Canny光伏组件故障定位方法。采用此方法可以实现从大面积光伏面板的红外图片中提取故障组件,将该方法与基于图像识别的故障诊断方法相结合能够大大提高光伏组件故障诊断的效率。首先利用改进遗传算法的Otsu算法实现光伏组件故障区域分割,再利用Canny算子实现故障边缘检测,最终完成故障定位。实验结果表明,与传统算法相比,所提算法故障定位快、准确率高,具有较好的实用性和应用价值。

In order to improve the efficiency of online fault diagnosis in photovoltaic systems,a fault location method for Otsu-Cannoy photovoltaic modules based on an improved genetic algorithm is proposed.Based on this method,it is possible to extract fault components from infrared images of large-area photovoltaic panels.Combining this method with image recognitionbased fault diagnosis methods can greatly improve the efficiency of photovoltaic module fault diagnosis.Firstly,the improved genetic algorithm Otsu algorithm is used to segment the fault area of photovoltaic modules,and then the Canny operator is used to achieve fault edge detection,ultimately completing fault localization.The experimental results show that compared with traditional algorithms,the proposed algorithm has fast fault location,high accuracy,and good practicality and application value.