摘要
为实现对钩舌的厚度、外涨、锁面磨耗量、销孔直径和锁面座入量的自动检测,提高货车钩舌的检测精度和效率,研究开发铁路货车钩舌自动检测系统。该系统主要由机械、控制和软件3个系统组成。采用基于视觉检测的线结构光传感器,并与运动机构相结合建立系统模型,并对系统模型参数进行标定。通过系统模型将机构的运动量和线结构光传感器扫描得到的检测数据转换为检测坐标系下的钩舌点云数据。采用范围过滤、均值滤波等算法对检测数据进行降噪处理,并通过特征提取算法计算检测值。实地试验性应用结果表明:该系统能够对所有型号货车钩舌的5个检测项目进行检测,达到了货车钩舌的检测要求。
An automatic detection system for railway freight car coupler knuckle was developed to achieve the automatic detection of the thickness,outward stretch,locking face abrasion loss,pin-hole diameter and locking face intake,as well as improve the detection accuracy and productiveness of freight car coupler knuckle.The system was mainly composed of mechanical system,control system and software system.Visual inspection based structured light sensor was adopted for the system,and the system model was built combining with movement mechanism.After calibrating the parameters of system model,the movement mechanism data and the detection data obtained by structured light sensor were transformed into the cloud data of coupler knuckle point under detection coordinate system via system model.Range filtering and mean filtering algorithms were adopted to denoise-process detection data,and feature extraction algorithm was used to calculate detection values.Field test application results show that the system can detect the five items of all types of freight car coupler knuckles,and satisfy the detection requirements for freight car coupler knuckle.
出处
《中国铁道科学》
EI
CAS
CSCD
北大核心
2013年第2期140-144,共5页
China Railway Science
基金
大连海事大学校青年创新基金资助项目(2011QN028)
关键词
货车钩舌
自动检测
检测模型
结构光传感器
扫描系统
参数标定
Freight car coupler knuckle
Automatic detection
Detection model
Structured light sensor
Scanning system
Parameter calibration
