基于轴线法与卡尔曼滤波的Baxter机器人标定
Kinematic Calibration Method of Baxter Robot Based on Screw-axis Measurement and Kalman Filter

为提高Baxter机器人的绝对定位精度，首先根据机器人操作系统（robot operating system, 简称ROS）的通用机器人模型统一机器人描述格式（unified robot description format,简称URDF）文件，详细分析了Baxter机器人的运动学模型，通过建立坐标间的约束条件进行模型简化，并提取待标定参数。针对机器人的多参数模型以及非几何参数影响较大的情况，提出一种结合轴线测量法与卡尔曼滤波系统辨识法的分步标定方法，分析了针对URDF模型的轴线法和卡尔曼滤波的标定原理，最后分别利用Matlab和激光跟踪仪进行仿真和实验验证。对Baxter机器人标定结果表明，该标定方法能够获得较准确的运动学几何参数，适用于ROS搭建的多几何参数的机器人及关节柔性较大的机械臂。
The kinematics model of the robot is analyzed according to the generic robot unified robot description format (URDF) file on robotic operating system(ROS). To simplify the kinematic model, some constraints are set for easily extracting of the calibrated parameters. Since the calibration model has multiple parameters and robot′s non-geometric parameters have poor effect on the robot′s accuracy, a method of fractional steps combined screw-axis measurement identification method and Kalman filter calibration method is proposed. In order to verify the feasibility of the method, simulation and experiment are made respectively through Matlab and laser tracker. The calibration results of baxter robot show that the method mentioned can obtain a more accurate kinematic geometry parameter. It can be applied to the robot built on ROS or ones with large joint compliance.