机器人视觉定位精度的灰色评定方法研究  

Location Accuracy Evaluation Method of Robot Vision System Based on Grey Generating Operation

作　　者：钱卫星[1] 黄丽亚[2] 

机构地区：[1]杭州职业技术学院电子与信息学院,杭州310018 [2]南京邮电大学电子科学与工程学院,南京210003

出　　处：《组合机床与自动化加工技术》2017年第6期30-33,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：浙江省教育厅科研项目(Y201327284);国家自然科学基金资助项目(61003237)

摘　　要：在不确定性动态扰动作用下,机器人视觉系统的各个位姿分量会产生不同量级的定位误差。为了量化各位姿分量的误差大小,提出一种视觉定位精度的灰色评定方法。首先,利用P3P特征点建模方法,建立机器人视觉系统的坐标映射模型;然后,利用GM(1,1)模型,分别对摄像机内参标定参数、2D坐标参数、3D坐标参数建立有干扰和无干扰条件下的灰色标准差模型;最后,将高斯噪音误差引入灰色标准差模型中,分析各视觉参数对机器人定位精度的量化影响。实验表明机器人视觉系统的位置精度优于±6mm,姿态精度优于±1.2°。The component both of the position and orientation has different error level under the effect of uncertain interference in practical engineering. To evaluate those different errors,a location accuracy analysis method is proposed based on Grey Generating Operation theory. Firstly,the coordinate mapping model of the robot vision system is built by the P3 P modeling method with three feature points. Secondly,the grey standard deviation of intrinsic parameters of camera,2D coordinate parameters,and 3D coordinate parameters are solved by GM（1,1） method both with interference and without interference. Thirdly,the grey standard deviation model of a robot vision system is motivated by Gaussian noise. The experimental result shows that the positional accuracy is ± 6mm,and the pose accuracy is ± 1. 2°,which shows the evaluating capability of our suggested method.

关 键 词：工业机器人 视觉系统 定位精度 灰色系统 

分 类 号：TH166[机械工程—机械制造及自动化] TG659[金属学及工艺—金属切削加工及机床]