基于机器视觉的水轮发电机上导主轴摆度超标自动检测方法
Automatic Detection Method for Excessive Swing of the Upper Guide Shaft of Hydroelectric Generator Based on Machine Vision
张佳辉1
作者信息
- 1. 国能陕西水电有限公司,陕西西安 710077
- 折叠
摘要
针对现有方法存在检测准确度低的问题,提出基于机器视觉的水轮发电机上导主轴摆度超标自动检测方法.先获取图像,运用拉普拉斯算子检测图像边缘,计算灰度图像中灰度值的频率.再运用特征匹配算法提取摆度图像中的关键特征点,利用Hough变换将水轮机上导主轴图像的直线特征映射到二维图像空间中,将直线特征转化为参数空间中的点.最后设定实际摆度过程中的阈值,通过将摆度值的计算结果与阈值进行比较,判断摆度是否超标.结果表明,10 次测试得到的实际值与测量值之间的误差在 0～0.1 mm之间,满足实际生产中对摆度测量的预期精度要求,应用效果较好.
Abstract
Aiming at the problems of low detection accuracy in existing methods,a machine vision based automatic detection method for excessive swing of the upper guide shaft of a hydroelectric generator is proposed.First,obtain the image,use Laplace operator for image edge detection,and calculate the frequency of grayscale values in the grayscale image.Then,feature matching algorithms are used to extract key feature points from the pendulum image,and the Hough transform is used to map the straight line features of the guide axis image on the water turbine to a two-dimensional image space,transforming the straight line features into points in the parameter space.Finally,set the threshold for the actual swing process,and compare the calculated swing value with the threshold to determine whether the swing exceeds the limit.The results show that the error between the actual value obtained from 10 tests and the measured value is between 0～0.1mm,which meets the expected accuracy requirements for swing measurement in actual production and has a good application effect.
关键词
机器视觉/水轮发电机/上导主轴/自动检测Key words
machine vision/hydroelectric generator/upper guide shaft/automatic detection引用本文复制引用
出版年
2024
NETL
NSTL
万方数据