Research On Catenary Geometric Parameters Detection Based On Point Cloud Data

As one of the most important mode of transportation in China,electrified railway plays an increasingly important role in China’s economic and social development.As an important part of the traction power supply system,the catenary has long been working in a complicated and harsh working environment and has no spare device.Therefore it has been regarded as the weakest part of the electrified railway power supply system.The geometric parameters of the catenary reflect working condition of the suspension and support devices,which affects the flow quality and driving safety of the train.At present,the detection of geometric parameters of catenary is mostly based on 2D images.Such detection methods are easily interfered by weather,2D image exposure,reflection on the surface of the object.Detection of dead spots may occur,affecting the final test results.With the advancement of image processing technology,especially the development of 3D image technology,detecting geometric parameters of catenary with higher precision and better detection efficiency in order to eliminate hidden dangers is very necessary.In this paper,3D point cloud processing methods are used to detect the geometric parameters of the catenary,and the detection method is verified by field experiment and computer simulation.This paper briefly introduces the process of deep camera acquiring 3D images and converting them into point cloud data before detecting geometric parameters.After obtaining the point cloud data of the catenary,corresponding methods are selected for data preprocessing,which is prepared for the subsequent geometric parameter detection.Then,for the detection of conductor height and stagger of the catenary,point cloud data of the contact lines is extracted and space continuous straight line was detected via the improved RANSAC method based catenary point cloud characteristics.According to the translation and rotation information,the relationship between the camera-coordinate system and the world-coordinate system is established,and the geometric parameters of catenary are calculated in accordance.The detection results are compared with the optical instrument measurements to verify the accuracy and effectiveness of this method.Finally,a geometric parameter detection method for catenary support and positioning device based on point cloud segmentation and corner detection is proposed.The point cloud segmentation is realized by R-RANSAC combined with Euclidean clustering.Multiple inclinations of supporting device are calculated according to the space vector information.Then,multiple corner points in the catenary supporting device are detected and located through corner detection algorithm based on 3D point cloud.After the above works,the inclination angles and the coordinates of the corner points are obtained,the structure of the catenary supporting device can be comprehensively analyzed.