首页|基于改进半径约束圆拟合方法的钢轨磨耗检测

基于改进半径约束圆拟合方法的钢轨磨耗检测

扫码查看
研究目的:钢轨磨耗检测是保证轨道质量,确保轨道系统安全运行的重要手段.传统钢轨检测方法依赖于人工使用特定的测量仪器来测量钢轨的轮廓,进而得到钢轨磨耗,耗费大量人力物力.基于结构光的非接触式测量方法可以快速、高效地对钢轨轮廓进行实时测量、磨耗计算,其检测效率高,安全性强,具有十分广阔的应用前景.研究结论:(1)利用离群点去除算法可有效地去除线结构光下采样得到的钢轨轮廓噪声;(2)利用插值算法结合60 kg钢轨自身结构比例特点可对钢轨轮廓进行精确分割;(3)采用基于半径约束的圆拟合算法提取测量件圆心坐标效果良好;(4)磨耗检测试验结果表明,采用本文的钢轨磨耗检测方法后,钢轨总磨耗均方根误差为0.08 mm,可满足钢轨磨耗检测的要求;(5)本研究成果可为轨道交通的运营维护提供一定参考.
A Rail Wear Detection Method with Improved Radius Constrained Circle Fitting Algorithm
Research purposes:Rail wear detection is an important method to ensure track quality and safe operation of the track system.Traditional rail inspection methods rely on the manual measuring of the rail profile with specific instruments and then the rail wear can be calculated.These methods require a great deal of manpower and material resources.The non-contact measurement method with line structured light provides quick and efficient measurement of the rail profile and wear calculation in real-time,and its high detection efficiency and safety have high potential applications.Research conclusions:(1)Outlier removal algorithm can effectively remove the rail profile noise due to sampling of the line structured light.(2)The interpolation algorithm combined with the characteristics of the structure proportion of 60 kg rail can accurately segment the rail profile.(3)The coordinates of the circle center of selected rail section is extracted precisely by radius constraint-based circle fitting algorithm.(4)The results of the wear detection experiment show that after using the rail wear detection method,the root mean square error of the total rail wear is 0.08 mm,which can meet the requirements of rail wear detection.(5)The research results can provide reference for the operation and maintenance of rail transit.

rail wear detectionline structured lightprofile matchingradius constraint

罗卿莉、陈志远、张书缤、石德斌、谭兆、甘俊

展开 >

天津大学精密测试技术及仪器全国重点实验室,天津 300072

中国铁路设计集团有限公司,天津 300308

钢轨磨耗检测 线结构光 轮廓匹配 半径约束

天津市轨道交通导航定位及时空大数据技术重点实验室开放基金天津市自然科学基金重点项目天津市交通运输科技发展计划天津市交通运输科技发展计划城市轨道交通数字化建设与测评技术国家工程实验室开放基金国家自然科学基金国家自然科学基金

TKL2023B1021JCZDJC006702020-022022-402021ZH044160144641801284

2024

铁道工程学报
中国铁道学会 中国铁路工程总公司 中国中铁股份有限公司

铁道工程学报

CSTPCD北大核心
影响因子:0.996
ISSN:1006-2106
年,卷(期):2024.41(3)
  • 8