基于机器视觉的塔式电站故障定日镜检测

Detect faulty heliostat in solar power plant based on machine vision

在塔式太阳能电站中,定日镜运行状态的稳定极大地影响电站的发电效率和运行安全。针对定日镜系统出现故障而导致定日镜无法正常转动的情况,采用机器视觉的方法对定日镜场进行检测。首先通过对定日镜的运动规律进行分析选择最佳检测时段,利用无人机对不同区域进行图像采集;然后对采集的图像进行滤波、级联边缘检测、矩形标记等一系列的图像处理操作,其中利用提出的级联边缘特征提取方法对物体在镜面的投影进行消除;最后利用故障诊断算法对采集的定日镜图像进行诊断,判断是否存在故障的定日镜并输出故障定日镜的位置。通过镜场模拟实验验证,该方法可以准确地识别出故障定日镜并显示出故障定日镜在镜场中的坐标。

In the solar power tower plant,the stability of the work of the heliostat greatly affects the power generation efficiency and operational safety of the power station.The machine vision method is applied to detect the broken heliostat due to malfunction of the heliostat system in the heliostat field.Firstly,selecting the best detection period by analyzing the motion of heliostats,the unmanned aerial vehicle(UAV) is used to collect images from different areas in the heliostat field.Then,a series of image processing operations are performed on the acquired image,including filter,edge detection and contour mark,among them the proposed cascaded edge feature extraction method is used to eliminate the shadows cast on the mirror surface by objects in this work.Finally,the fault diagnosis algorithm is served to discriminate the acquired heliostat image where whether there is a faulty heliostat and outputs the position of the fault heliostat.It is verified that the method can accurately identify the fault heliostat and display the coordinates of the fault heliostat by the heliostat field simulation experiment.