【作者】 王健；

【导师】 罗隆福；

【作者基本信息】 湖南大学 ， 电气工程， 2021， 硕士

【摘要】 高速铁路接触网在高铁供电系统中有着巨大的作用,而且运行线路上没有备用装置,高速列车行驶带来的剧烈震动和恶劣天气也给接触网的安全运行带来了隐患,因此保障高铁供电安全就需要对接触网进行及时有效的监测和检测。目前铁路上对接触网的检测方法主要有人工巡检方式、工作人员人工查看图像方式和应用计算机视觉算法对图像进行分析。采用图像处理算法逐渐成为一种高效的接触网检测方式,但是仍然存在很大的提升空间。本文在研究分析了接触网零部件结构分布以及目前对于接触网零部件缺陷检测算法的进展的基础上,针对高速铁路接触网紧固设备开口销存在缺失、松脱、安装不规范等故障状态,提出一种结合目标检测算法,语义分割算法和形态学处理方法的三阶段高速铁路接触网开口销缺陷检测算法,并对多种环境下的接触网图片进行了测试,结果表明了本文提出算法的有效性。首先,基于YOLOV4算法采用两次定位方法定位开口销,先通过初定位完成高铁接触网中的五个关键连接点的定位,然后通过二次定位完成开口销的精定位。解决了因开口销在接触网图片中占比小导致定位不准的问题。然后采用Deeplabv3+算法对开口销进行语义分割,语义分割算法作为深度学习算法的一种,可以有效的区分物体类别,并形成相应的语义区域,本文采取语义分割算法主要是用于提取接触网中的开口销语义信息。解决了背景噪声影响开口销缺陷检测准确率的问题。最后基于开口销的语义分割图片进行分类,分类依据为开口销的头部,躯干和尾部的语义信息。为了验证本文提出方法的适应性和准确性,对多条线路,多种环境下的接触网图片进行测试。同时,与其他深度学习算法进行比较。实验结果表明,本文所采用的方法可以较准确的检测出缺失和松脱的开口销,保障接触网的稳定运行。同时也表明应用深度学习算法可以更准确,更高效的完成接触网检测,因此也可以应用到接触网其他零部件的缺陷检测上。更多还原

【Abstract】 The catenary structure of high-speed railway plays an important role in the power supply system of high-speed railway.Moreover,there is no standby device on the running line,and the violent vibration brought by the high-speed train and bad weather also bring hidden dangers to the safe operation of the catenary.Therefore,to ensure the safety of high-speed rail power supply,timely and effective monitoring and testing of catenary is needed.At present,the detection methods of catenary on railway mainly include manual inspection,manual viewing of images by workers and analysis of images by computer vision algorithm.Using image processing algorithm has gradually become an efficient catenary detection method,but there is still much room for improvement.Aiming at the fault states of split pins in the catenary fastening devices of highspeed railway,such as missing,loosening and improper installation,a three-stage method for detecting defects of split pins is proposed based on the research and analysis of the structural distribution of catenary components and the current progress of defect detection algorithms for catenary components.This method is based on object detection algorithm,semantic segmentation algorithm and morphological processing method.And the effectiveness of the algorithm proposed in this paper is verified by experimental tests.Firstly,split pins are localized by a two-stage positioning method based on the YOLOV4 algorithm.The first stage is used to localize five joint components on catenary support devices and the second stage is applied to locate the split pins in the joint component images.In this way,the problem of inaccurate positioning due to the small proportion of split pins in the catenary picture is solved.Then,the deeplabv3+ algorithm is implemented for semantic segmentation on split pin images.As a kind of deep learning algorithm,semantic segmentation algorithm can effectively distinguish object categories and form corresponding semantic regions.The semantic segmentation algorithm adopted in this paper is mainly used to extract the semantic information of split pins in catenary.Finally,split pins are classified based on semantic segmentation images,in accordance with split pins’ semantic information of both head,body and tail.In order to verify the adaptability and accuracy of the proposed method,the catenary support devices images in multi high-speed railway lines and multi environments were tested.Meanwhile,our algorithm is compared with other deep learning algorithm.The results show that the proposed method has a higher accuracy in detecting defects of split pins,which can guarantee the stable operation of catenary support devices.At the same time,it also shows that the application of deep learning algorithm can complete the inspection of catenary more accurately and efficiently,so it can also be applied to the defect inspection of other parts of catenary.更多还原