摘要
受电弓的异常状态是对高速铁路运营安全影响较大且备受关注的问题。基于计算机视觉的受电弓滑板缺陷智能检测技术,结合改进的YOLOv4模型与边缘提取等传统图像处理算法,研究适用于受电弓滑板监测装置(5C)的缺陷智能识别模型。铁路现场试验证明该智能识别模型在受电弓滑板缺陷检测中的有效性和实时性。
The abnormal condition of the pantograph is of great influence on the operation safety of high speed railways and has got much attention. Based on the computer vision intelligent defect-detection technology of pantograph pans, combined with improved traditional image processing algorithms such as YOLOv4 model and edge extraction, this paper makes a study of the intelligent defect-recognition model applicable to the monitoring device(5 C) of the pantograph pan. The effectiveness and real-time performance of this intelligent recognition model in the detection of the pantograph pan defects are demonstrated in the railway field tests.
作者
莫小凡
王科理
潘长清
赵文军
占栋
MO Xiaofan;WANG Keli;PAN Changqing;ZHAO Wenjun;ZHAN Dong(Standards&Metrology Research Institute,China Academy of Railway Sciences Corporation Limited,Beijing 100081,China;China Railway Test&Certification Center Limited,Beijing 100081,China;Chengdu Tangyuan Electric Co ltd,Chengdu Sichuan 610000,China)
出处
《中国铁路》
2022年第2期148-155,共8页
China Railway
关键词
受电弓
受电弓滑板监测装置(5C)
智能识别
图像处理
pantograph
monitoring device(5C)of pantograph pans
intelligent recognition
image processing
