基于数字散斑的轮轨垂向载荷识别方法研究

Research on wheel-rail vertical load detection based on digital speckle pattern

快捷有效地进行轮轨垂向载荷检测对保障轨道车辆的服役可靠性与运行安全具有重要意义。现有轮轨载荷检测方法尽管能实现轮轨垂向载荷的有效检测,但其测量精度受传感器布设方案、系统标定的限制,仅能实现定点检测。因此,引入数字散斑图像相关技术,为轮轨垂向载荷的快速、非定点检测提供新的技术方案。为了实现钢轨轨侧应力场的非接触、快速测量,通过左右相机同步采集垂向载荷作用下的钢轨散斑图像序列,并通过图像相关理论模型进行变形前后子区像素点的匹配和应力场计算。为了利用数字散斑检测获得的钢轨轨侧应力分布进行轮轨垂向载荷的识别,提出一种基于极限学习机(ELM)的轮轨垂向载荷识别方法。该方法采用Workbench有限元软件建立钢轨的数值模型,通过数值模拟结果选取应力变化相对敏感的区域作为轮轨垂向载荷检测的兴趣域,基于应力场数据集和对应的载荷数据集设计ELM网络参数,进而实现轮轨垂向载荷的自动识别。为了验证提出的轮轨垂向载荷识别模型的有效性,搭建基于XT-DIC的轮轨垂向载荷数字散斑检测实验平台。实验结果表明,用Y和Z方向模应力输入的方式构建的ELM模型有着最佳的识别性能,其垂向载荷预测误差仅有5.357%。研究结果为轮轨垂向载荷的非定点快速检测提供了新的可靠途径。

Fast and effective detection of wheel/rail vertical load is of great significance for ensuring the reliability and safety of rail vehicles in service.Although existing wheel/rail load detection methods can effectively detect wheel/rail vertical loads,their measurement accuracy is limited by sensor deployment schemes and system calibration,and can only achieve fixed point detection.Therefore,this study introduced digital speckle image correlation technology to provide a new technical solution for rapid and non-fixed point detection of wheel/rail vertical loads.In order to achieve non-contact and rapid measurement of the stress field on the rail side,a sequence of rail speckle images under vertical load was synchronously collected by left and right cameras,and the sub-region pixel point matching and stress field calculation before and after deformation were carried out through image related theoretical models.In order to identify the wheel/rail vertical load using the rail side stress distribution obtained by digital speckle detection,a wheel/rail vertical load identification method based on the Extreme Learning Machine(ELM)was proposed.This method used Workbench finite element software to establish a numerical model of the steel rail.Through the numerical simulation results,areas that are relatively sensitive to stress changes were selected as the interest domain for wheel/rail vertical load detection.Based on the stress field dataset and corresponding load dataset,ELM network parameters were designed to achieve automatic recognition of wheel/rail vertical load.To verify the effectiveness of the proposed wheel/rail vertical load identification model,a digital speckle detection experimental platform for wheel/rail vertical load based on XT-DIC was established.The experimental results show that the ELM model constructed by inputting mode stress in the Y and Z directions has the best recognition performance,with a vertical load prediction error of only 5.357%.The research results provide a new and reliable way approach for non fixed point rapid detection of wheel/rail vertical load.