钢轨焊前断面廓形参数检测方法研究  

Study on Methods for Detecting Pre-welding Rail Profile Parameters

作　　者：谢雯 李存荣[1] XIE Wen;LI Cun-rong(School of Mechanical and Electronic Engineering,Wuhan University of Technology,Wuhan 430 0 7 0,Chin)

机构地区：[1]武汉理工大学机电工程学院,武汉430070

出　　处：《铁道标准设计》2018年第4期40-46,共7页Railway Standard Design

基　　金：国家自然科学基金课题(71171154);湖北省科技支撑计划项目(2015BAA063)

摘　　要：钢轨断面的廓形参数对钢轨焊接后的平直度、焊后打磨、钢轨品质等影响重大。针对钢轨焊接前的轨端廓形检测,在激光传感器采集的廓形数据基础上,详细论述数据的处理方法和过程,主要包括数据的预处理、数据的融合、分段拟合、廓形的计算等,分析不同方法针对该研究目的的可行性、适应性以及准确性,最终选择移动平均法进行数据的预处理,采用矩形标准块和圆柱标准块标定进行数据融合,基于协方差矩阵的特征值选取分段点,最小二乘法进行分段拟合。试验数据表明,该方法对数据处理的效果良好,可以高效精确地检测出钢轨焊前断面的廓形尺寸。The rail profile parameters determine the smoothness,the post-welding grinding and the final quality of the rail. Based on the profile data collected by laser sensors,data processing methods are discussed in detail,including data fusion,segmentation fitting,contour calculation and so on. This paper analyzes the feasibility,accuracy and adaptability of different methods,conducts data preprocessing by moving average method and data fusion by the rectangular block and cylinder block calibration.Segmentation points are selected based on covariance matrix and segmentation fitting is completed by least square method. The experimental data shows that the method has good effect on data processing,and can efficiently and accurately detect pre-welding rail profile.

关 键 词：钢轨 焊前检测 钢轨廓形 激光传感器 数据处理 

分 类 号：U213.4[交通运输工程—道路与铁道工程]