多传感信息融合机器人标定方法  

Robot calibration method based on multi-sensor information fusion

作　　者：张湧涛 金爽 王一 

机构地区：[1]华北理工大学电气工程学院,河北唐山063009

出　　处：《中国测试》2017年第4期56-62,共7页China Measurement & Test

基　　金：国家自然科学基金青年科学基金项目(51505125)

摘　　要：该文提出一种基于多传感信息融合技术进行机器人运动学标定的方法。首先通过对机器人的指数积(POE)正向运动学模型取微分的方式建立末端执行器的误差模型,利用ROMER-RA7520绝对关节臂测量机和安装在末端工具上的多个传感器分别采集末端执行器的位置和姿态信息;随后根据坐标统一原理将测量机的位置测量数据和多传感器的姿态量测数据转换到机器人基础坐标系下,实现位姿数据的空间配准;接着运用自适应加权融合算法融合处理经过空间配准后的位姿数据,得到末端执行器位姿测量值;最后应用迭代最小二乘法求解出参数偏差。KR5arc机器人标定仿真实验表明:该方法可大幅度提升机器人在任意位姿下的定位、定姿精度。A method of robot kinematics calibration based on multi-sensor information fusion technology is proposed. Firstly, an error model of end effeetor is established by differentiating the POE forward kinematics model of robot, and then ROMER-RA7520 absolute articulated arm measuring machine and many sensors mounted at the end tool are used to collect the end effector position and attitude information. After that, the position measurement data of the measuring machine and the attitude measurement data of multi-sensor will be converted to the robot base coordinate system according to unity principle of the coordinate to realize the special alignment of position and attitude data. And then, the adaptive weighted fusion algorithm is used to have fusion processing of the position and attitude data after special alignment and get the measured position and attitude data of end effector. Finally, iterative least square method is used to get the parameter deviation. KR5arc robot calibration simulation experiment shows that the method can greatly improve the positioning and attitude determination accuracy of the robot under any position and attitude.

关 键 词：指数积 测量臂 空间配准 多传感信息融合 参数标定 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]