基于多点动态捕捉的机器人关节位移与几何参数标定

机器人关节位移及几何参数误差是导致位姿出现误差的主要原因，对二者进行高精度标定可提高机器人的绝对精度．传统标定方法的精度受机器人位姿的影响，且需多次实验．本文提出一种基于多点动态捕捉测量的标定方法，可通过单次实验，同时标定串联机器人的关节位移和几何参数．首先基于多点动捕测量获取机器人各连杆上靶标点的3D坐标，结合罗德里格斯变换实现关节位移的求解．其次结合机器人运动学模型和坐标转换关系，实现几何参数误差小量的标定．对方程线性化处理以提高计算效率，并用最小二乘法降低噪声对结果的影响．最后通过对6自由度串联机器人进行仿真，验证了该标定方法的可行性．

A Calibration Method of Robot Joint Displacement and Geometric Parameters Based on Multi-Point Motion Capture System

Robot joint displacement error and geometric parameter error are the main causes of pose deviation. The absolute accuracy of serial robot can be improved by high-precision calibration of them. However, the accuracy of traditional calibration methods is affected by the pose of the robot, and repeated experiments are usually needed. In this paper, we proposed a calibration method based on multi-point motion capture measurement, which could calibrate the joint displacement and geometric parameters of the serial robot through a single experiment. Firstly, the 3D coordinates of the markers on each robot link were obtained based on multi-point motion capture measurement, and the joint displacement was obtained by Rodrigues transform. Secondly, the robot kinematics model and coordinate transformation relationship were combined to calibrate the error of geometric parameters. The equations were linearized to improve the calculation efficiency, and the influence of noise on the results was reduced by the least square method. Finally, the feasibility of the calibration method was verified by the simulation of a 6-DOF serial robot.