Fuzzy Adaptive Model Predictive Control for Image-based Visual Servoing of Robot Manipulators With Kinematic Constraints

This paper presents a novel image-based visual servoing (IBVS) controller for a six-degree-of-freedom (6-DoF) robot manipulator by employing a fuzzy adaptive model predictive control (FAMPC) approach. The control strategy allows the robot to track the desired feature points adaptively and fulfill kinematic constraints appearing in a vision-guided task with different initial Cartesian poses. To this aim, the successive linearization method is firstly employed to transform the nonlinear IBVS model to the linear time-invariant (LTI) one at each sampling instant. The nonlinear optimization problem is therefore degraded into a convex quadratic programming (QP) problem. Subsequently, a fuzzy logic is exploited to tune the weighting coefficients in the cost function on the basis of image pixels changes at each step, endowing the reliable adaptation capabilities to different working environments. Experimental comparison tests performed on a 6-DoF robot manipulator with an eye-in-hand configuration are provided to demonstrate the efficacy of the proposed controller.