Image Based Visual Servoing Model Predictive Control of Manipulator

Visual servoing as an important implementation of machine intelligence has been widely studied. While in the visual servoing control task of multi-joint manipulator, the task often fails due to the occurrence of camera retreat phenomenon. The present paper considers the combinations of the image based visual servoing and model predictive control (MPC) scheme to control manipulator. The visual servoing control of the manipulator system includes the object detection system, visual servoing controller and manipulator. By considering the MPC scheme with dynamic weight matrices, the phenomenon of camera retreat in the process of manipulator visual servoing is properly dealt with. The designed methods are verified in the simulation experiments.