摘要
为了解决机器人在生产应用中的精度问题,提出一种基于视觉测量方法的运动学标定,提高机器人末端位置精度。根据该Tripod机器人的结构特点及几何参数,建立运动学模型;分析机构多项几何误差源,采用解析法得到动平台末端位置与机器人几何误差源间的误差映射模型,进一步简化得到零点误差映射模型;提出一种采用视觉作为误差测量的测量方法,并采用最小二乘法构造辨识模型对零点误差进行辨识;对标定实验进行测点规划,通过拟定的实验步骤对机器人进行标定实验;用修正输入的补偿模型对误差进行补偿,并通过激光跟踪仪验证标定前后的位置精度。经验证末端位置精度显著提高,表明了该视觉测量方法的可行性和有效性,以及该测量方法的简单便捷。
In order to solve the precision problem of robot in productive application,a kinematic calibration based on vision measurement method is proposed to improve the precision of endposition of robots.Kinematics model is established according to structural characteristics and geometric parameters of this Tripod robot.With analyzing on geometric error sources of the mechanism,the error mapping model between the end-position of the mobile platform and the geometric error sources of the robot is obtained by analytical method,and the zero point error mapping model is further simplified.A measurement method using vision as error measurement is proposed,and the zero point error is identified with the least squares structure identification model.The calibration experiment is carried out on the robot through the proposed experimental steps with measure-point layout.And the error is compensated by the compensation model of the modified input,and the position precision before and after calibration is verified by the laser tracker.The precision of the end-position is verified to be significantly improved,that shows the feasibility and effectiveness of the visual measurement method on kinematic calibration,and the simplicity and convenience of the measurement method.
作者
杨继东
陶金池
覃畅
YANG Jidong;TAO Jinchi;QIN Chang(College of Mechanical Engineering,Chongqing University,Chongqing 400044,China)
出处
《传感器与微系统》
CSCD
北大核心
2021年第5期25-29,共5页
Transducer and Microsystem Technologies
基金
国家重点研发计划资助项目(2016YFC0803200)。
