一种结合手眼标定与TCP标定的标定方法OA
A calibration method combining hand-eye calibration and TCP calibration
传统机器人工具中心点(Tool Centre Point,TCP)标定和手眼标定分开进行,会产生较大的累积误差.提出一种基于 TCP 标定与手眼标定协同标定方法.采用九点法及平面 TCP 标定法获取手眼标定初始矩阵及 TCP 初始偏移量,并建立基坐标系下空间任意点的坐标方程.通过李代数线性降维方法获取 TCP 值、旋转角、相机高度及手眼矩阵四个误差补偿量,并得到手眼标定与 TCP 标定的精确值.仿真与实验结果表明,采用的协同标定算法的平均标定位置误差比分离标定法降低了约34%,提升了机器人的标定精度.通过实验验证表面,本算法减小了位置误差与姿态误差,提升了机器人的标定精度,可以满足更多的工作场景需求.
Traditional methods of calibrating the tool center point(TCP)and conducting hand-eye calibration were performed separately,resulting in significant cumulative errors.This paper proposed a collaborative calibration method combining TCP calibration and hand-eye calibration.The nine-point method and planar TCP calibration method were adopted to obtain the initial hand-eye calibration matrix and initial TCP offset,and the coordinate equations for arbitrary points in space were established in the base coordinate system.With the Lie algebra linear dimensionality reduction method,TCP values,rotation angles,camera height,and error compensation quantities of the hand-eye matrix were acquired,and the precise values for hand-eye calibration and TCP calibration were derived.Simulation and experimental results demonstrated that the average calibration position error of the proposed collaborative calibration algorithm was approximately 34%lower than that of the separate calibration method,thus improving the robot's calibration accuracy.
袁彬;丁伟;徐子翔;张润梅;陈中;李佳祥
安徽建筑大学 建筑机械智能制造重点实验室,合肥 230601安徽建筑大学 建筑机械智能制造重点实验室,合肥 230601安徽建筑大学 建筑机械智能制造重点实验室,合肥 230601安徽建筑大学 建筑机械智能制造重点实验室,合肥 230601安徽建筑大学 建筑机械智能制造重点实验室,合肥 230601安徽建筑大学 建筑机械智能制造重点实验室,合肥 230601
信息技术与安全科学
手眼标定TCP标定九点标定协同标定李代数误差补偿
hand-eye calibrationTCP calibrationnine-point calibrationcollaborative calibrationLie algebraserror compensation
《哈尔滨商业大学学报(自然科学版)》 2026 (4)
450-456,7
安徽建筑大学项目(JZ202364)安徽建筑大学项目(JZ202391)国家自然科学基金(52275483)工程机械智能制造重点实验室开放基金(IMCM2023KF02)陕西省公路工程机械重点实验室开放基金(300102252507)安徽省仿真设计和现代制造工程技术研究中心开放基金(SGCZZZD2101)
评论