基于图像的直接视觉伺服控制器设计与仿真

对于手眼机器人视觉伺服,考虑机器人的运动学及动力学特性,在物体的深度未知及机器人的基坐标系与物体的坐标系之间的相似变换矩阵未定标的情况下,设计了直接视觉伺服控制器,利用自适应算法对深度及上述相似变换矩阵的参数进行在线估计,实现了基于图像误差的定位控制,在系统的工作空间范围内,闭环系统是渐近可稳定的。该方法不需要物体的几何模型及深度的精确值,仿真结果验证了其有效性。

Image-Based Direct Visual Servo Controller Design and Simulation

Considering robot kinematics and dynamics for visual servoing with eye-in-hand configuration, a direct visual servo controller was proposed to regulate a set of feature points in the image plane to desired positions in the case of unknown object depth and the homogeneous transformation matrix between the robot frame and the object frame. One adaptive algorithm was designed to estimate the depth of selected features of the object and another to estimate the unknown homogeneous transformation matrix on-line. By invoking Lyapunov direct method, it is proved that the overall closed-loop system is locally stable. The method does not need knowledge of geometrical three-dimensional model of the object, A simulation is carried out to show its validity.