Generalized Predictive Control of an anthropomorphic robot arm for trajectory tracking

Abstract

This paper presents an effective model-based predictive approach for the precise trajectory tracking of an anthropomorphic robot arm. The proposed control strategy is based on feedback linearization and linear Generalized Predictive Control, requiring no on-line optimization procedure. Experimental evaluation of the proposed method and its comparison with two classic robot control approaches illustrate its tracking performances and robustness with respect to non-compensated load variations.