Position-based visual servoing of manipulator using model predictive control

In order to solve the problem of acquiring the position of the target and tracking control of the target, this paper designs a visual servo system for manipulator position based on model predictive control. First, the initial camera position value is obtained by solving the relative position relationship between the camera and the target by the Perspective-n-Point algorithm. Second, the Levenberg–Marquardt algorithm is adopted to reduce the influence of noise and other factors on measurement accuracy. It is adopted to optimize the camera position by minimizing the reprojection error. Third, considering the motion space constraints during the movement of the manipulator, the model predictive control method is employed to solve the visual servo control problem with constraints. The monocular vision system is used to measure the position of the end of the manipulator and the target to provide feedback for the model predictive control. Finally, the simulation experiment is provided to verify the validity and effectiveness of the proposed method.