刘运毅, 黎相成, 黄约, 唐明福, 秦德茂, 农真. 基于极大似然估计的工业机器人腕部6维力传感器在线标定[J]. 机器人, 2019, 41(2): 216-221,231. DOI: 10.13973/j.cnki.robot.180203
引用本文: 刘运毅, 黎相成, 黄约, 唐明福, 秦德茂, 农真. 基于极大似然估计的工业机器人腕部6维力传感器在线标定[J]. 机器人, 2019, 41(2): 216-221,231. DOI: 10.13973/j.cnki.robot.180203
LIU Yunyi, LI Xiangcheng, HUANG Yue, TANG Mingfu, QIN Demao, NONG Zhen. Online Calibration for the 6-axis Force Sensor in the Wrist of Industrial Robot Based on Maximum Likelihood Estimation[J]. ROBOT, 2019, 41(2): 216-221,231. DOI: 10.13973/j.cnki.robot.180203
Citation: LIU Yunyi, LI Xiangcheng, HUANG Yue, TANG Mingfu, QIN Demao, NONG Zhen. Online Calibration for the 6-axis Force Sensor in the Wrist of Industrial Robot Based on Maximum Likelihood Estimation[J]. ROBOT, 2019, 41(2): 216-221,231. DOI: 10.13973/j.cnki.robot.180203

基于极大似然估计的工业机器人腕部6维力传感器在线标定

Online Calibration for the 6-axis Force Sensor in the Wrist of Industrial Robot Based on Maximum Likelihood Estimation

  • 摘要: 针对工业机器人实时监测腕部工具工作受力情况的需求,提出了基于极大似然估计的在线标定算法.首先,在机器人末端工具的活动负载部分安装6维力传感器,并实时采集力与力矩数据以及机器人工具运动轨迹.然后,考虑到不同运行速度下运动振动的干扰,根据系统力学关系求解工具重心坐标、机器人安装倾角、力传感器零点以及负载重力.最后,在10 mm/s~1000 mm/s不同速度条件下进行实验,分析求解结果的一致性,并与最小二乘法进行对比.对比结果显示基于极大似然估计可以把工具重心平均标准差从0.67 mm降到0.23 mm,力零点标准差从0.73 N降到0.27 N,力矩零点标准差从0.29 N.m降到0.05 N.m.实验结果表明极大似然估计能有效抵抗机器人高速运动引起的干扰,可以应用于机器人高速运动情况下的实时在线零点标定.

     

    Abstract: An online calibration algorithm based on the maximum likelihood estimation is proposed to monitor the forces on wrist tools in real-time for industrial robots. Firstly, the 6-axis force sensor is installed in the robot end tools to collect the force, the torque and the motion path of the robot tool in real-time. Then, the gravity coordinate of the tool, the installation angle of the robot, the zero offset of the force sensor and the gravity of the load are calculated according to the force relation of the system while considering the motion vibration interferences at different speeds. Finally, the experiments at different speeds between 10 mm/s~1000 mm/s are conducted, and the consistence of the solution results are analyzed and compared with the results of the least square method. The comparison results show that the maximum likelihood estimation method reduces the average standard deviation of the gravity center from 0.67 mm to 0.23 mm, the standard deviation of the force zeros from 0.73 N to 0.27 N, and the standard deviation of the torque zeros from 0.29 N.m to 0.05 N.m. The experimental results show that the maximum likelihood estimation method can effectively resist the interference caused by high-speed motion of the robot, and can be applied to real-time and online zero calibration of the robot in the case of high-speed motion.

     

/

返回文章
返回