This paper presents a predictive visual servo kinematic control scheme for a robotic manipulator with the eye-in hand configuration to perform autonomous capture of a non-cooperative space target with unknown motion. The unknown motion is estimated by an integrated algorithm of the photogrammetry and adaptive extended Kalman filter, in which the eye-in-hand configuration of the vision system improves the accuracy of the motion estimation as it approaches the target. Based on the vision feedback, a dynamic trajectory of robotic manipulator is planned in real time at each sampling instant and the end-effector of the manipulator moves towards the predicted position of the target at the next time instant incrementally. In this way, the multiple solutions problem of inverse kinematics in the joint space is effectively avoided and the robotic manipulator could intercept the noncooperative target for a fast rendezvous. Validation experiments are performed on a custom built robotic manipulator with an eye-in-hand configuration. The experimental results demonstrate the effectiveness and robustness of the proposed control scheme.
展开▼
