白车身曲面轮廓度误差评价方案研究  

Research On Error Evaluation Scheme of Body In White Surface Profile

作　　者：陆雨薇 刘宇鹏 郑伟 王柏杰 Lu Yu-wei;Liu Yu-peng;Zheng Wei;Wang Bo-jie(Guangxi Collaborative Innovation Center of Earthmoving Machinery,School of Mechanical and Automotive Engineering,Guangxi University of Science and Technology,Guangxi Liuzhou 545006,China;Guangxi New Energy Vehicle Laboratory,Guangxi Liuzhou 545007,China;Liuzhou Huxin Automobile Technology Co.Ltd,Guangxi Liuzhou 545616,China)

机构地区：[1]广西科技大学机械与汽车工程学院广西土方机械协同创新中心,广西柳州545006 [2]广西新能源汽车实验室,广西柳州545007 [3]柳州沪信汽车科技有限公司,广西柳州545006

出　　处：《内燃机与配件》2024年第13期14-17,共4页Internal Combustion Engine & Parts

基　　金：广西科技计划(桂科AD22080042);中央引导地方科技发展资金项目(桂科ZY21195005)。

摘　　要：针对三坐标测量机测量点云数据过程中存在点云坐标系不对应问题,本研究提出改进ICP点云配准方法,对理论与实测点云进行配准。结合极限逼近的思想,将点云轮廓度转化为局部点云平面度的误差评价过程,利用最小区域法评价点云轮廓度误差。经对仿真实验分析结构验证,改进的点云配准方法可收敛于更优的目标函数,配准效果分别提升23%、20%。评价方案所得到的轮廓度误差减少了30%,有效提高轮廓度评价精度,提高工件的合格率。Aiming at the problem that point cloud coordinate system does not correspond in the process of measuring point cloud data by CMM,this study proposes an improved ICP point cloud registration method to register theoretical and measured point clouds.Combining with the idea of limit approximation,the point cloud profile is transformed into the error evaluation process of local point cloud flatness,and the minimum region method is used to evaluate the point cloud profile error.According to the analysis structure of the simulation experiment,the improved point cloud registration method can converge to the better objective function,and the registration effect is improved by 23% and 20% respectively.The profile error obtained by the evaluation scheme is reduced by 30%,which effectively improves the accuracy of profile evaluation and the pass rate of workpiece.

关 键 词：白车身型面 点云配准 轮廓度误差评价 算法设计 

分 类 号：U466[机械工程—车辆工程]