Model free position-force control of environmental constrained reconfigurable manipulators based on adaptive dynamic programming

This article proposes a model free position-force control method for uncertain environmental constrained reconfigurable manipulators based on adaptive dynamic programming algorithm. Through the analysis of kinematic uncertainties, an adaptive estimation scheme is designed to obtain the approximate contacted torque. The contacted torque is generated due to the interaction of the manipulator's end-effector with the uncertain environment. Then, the performance index function is defined by utilizing the joint position, contacted torque tracking errors and uncertain environmental factors. The presented neural network-based observer is utilized to learn the dynamic model. On the basis of policy iteration algorithm, the corresponding Hamiltonian-Jacobi-Bellman equation is addressing by employing the critic NN structure. Thus, the model free position-force control strategy is obtained. Based on Lyapunov stability theorem, the tracking error of the reconfigurable manipulator is proved to be ultimately uniformly bounded. Eventually, simulations and experiments are illustrated the effectiveness of the developed controller.