基于极小极大算法的双臂机器人精度提升方法

Minimax algorithm for improving the accuracy of dual-arm robots

针对双臂机器人绝对精度和协作精度的指标要求,基于激光跟踪仪测量的双臂末端三靶球点位数据,使用极小极大的搜索算法,提出了鲁棒的双臂机器人运动学几何参数校准方法。该方法分两步:首先,分别对左右两臂的几何参数进行校准,在左(右)臂末端安装三个靶球,以三个测量点中的最差绝对定位误差为指标函数建立非线性优化问题,确保校准结果的鲁棒性并提高机器人末端的定向精度,采用二次序列规划法求解;然后,对左右两臂间的坐标准直进行校准,以三靶球中左右臂间的最差协作定位误差为指标函数,采用极小极大算法提升双臂间的协作位姿精度。

Aiming at the requirements of the absolute accuracy and cooperation accuracy of a dual-arm robot,based on the data of the three-target ball position measured by a laser tracker,and by using a minimax search algorithm,a robust calibration method for kinematic parameters of the dual-arm robot was proposed.The method was divided into two steps as follows.Firstly,the geometric parameters of the left and right arms were calibrated separately.Three target balls were installed at the end of the left(right)arm,and a nonlinear optimization problem was established,with the worst absolute positioning error of the three measurement points was used as an index function,to ensure the robustness of the calibration process and also to improve the orientation accuracy of the end-point.The quadratic sequence programming method was used to solve the nonlinear optimization problem with inequality constraints.Secondly,the transformation between the left and right arms was identified.Taking the worst collaborative positioning error between the left and right arms in the three target balls as the target function,the minimax algorithm was used to improve the cooperative pose accuracy between the two arms.