Predictive image based visual servoing of manipulators with multiple constraints

Abstract

A novel image based visual servoing (IBVS) method under the predictive control frame is proposed in order to solve the visual tracking problem of robot manipulator with multiple constraints. A vision predictive frame is presented by employing the motion model between camera's velocity and image features. And the design step of vision predictive controller is presented. Then, an effective vision predictive control method with a suitable predictive time horizon is obtained. In simulation, the role of predictive time horizon is analyzed and evaluated, and several simulation results are gained by using the proposed method. Finally, the interesting and satisfactory results are illustrated.