Asymptotically stable adaptive robot control exploiting the essential parameter set

Abstract

The authors introduce a procedure to determine the set of essential parameters of a robot manipulator dynamic model. Then they use the Lyapunov stability theory to construct an adaptive control algorithm based on the model containing only the parameters that belong to the determined set. The proposed control algorithm is computationally efficient and it requires neither inversion of the manipulator inertial matrix nor joint acceleration measurement. Asymptotic stability of the proposed control algorithm is proved.<>