基于多视觉传感器的受电弓滑板磨耗检测系统设计

Design of Wear Detection System of Pantograph Slide Plate Based on Multi-vision Sensor

针对列车受电弓滑板磨耗检测需求的不断增加,设计一种基于多视觉传感器的受电弓滑板磨耗自动检测系统。该系统通过安装在检测线龙门架上的多台视觉传感器获取受电弓滑板高精度三维点云,通过建立基于多视觉传感器的组合标定模型,将各视觉传感器坐标系下的滑板数据统一于全局坐标系下,拼接得到完整的滑板表面轮廓;采用基于矩形包围盒的体素栅格对原始滑板点云进行下采样,采样后的点云通过密度聚类算法对滑板和托板进行目标数据聚类;最后拟合托板平面,根据滑板平面和托板平面的高度差计算滑板剩余磨耗量。通过试验证明,该系统在保持较高测量精度的情况下,同时具有较高重复精度、稳定可靠,达到工程实际应用要求。

In view of the increasing demand for the wear detection of pantograph slide of train,an automatic wear detection system of pantograph slide was designed based on multiple vision sensors.This system obtained the pantograph slide high precision 3D point cloud through multiple vision sensors installed on the detection line gantry frame.By establishing the combined calibration model based on multiple vision sensors,the slide data in the coordinate system of each vision sensor was unified in the global coordinate system,and the complete surface contour of the slide was obtained by splicing.The voxel grid based on rectangular enveloping box was used to carry out down sampling to the original skateboard point cloud,and the sampled point cloud was used to cluster the target data of skateboard and pallet through density clustering algorithm.Finally,the supporting plate plane was fitted and the residual wear of the sliding plate was calculated according to the height difference between the sliding plate plane and the supporting plate plane.The experiment shows that the system has high repeatability,stability and reliability,and meets the requirements of practical application under the condition of maintaining high measurement accuracy.