A robust force observer for robot manipulators subjected to external disturbance

The force estimation problem in robotic systems is studied in this paper. To isolate the effects of external disturbance on force estimation, a neural network based nonlinear disturbance observer is proposed. Hence, unlike existing force observers in the literature, force estimation in the presence of external disturbances is guaranteed. Based on the Lyapunov stability criterion, sufficient conditions to guarantee the stability of the proposed observer are derived. Simulation results performed on a two link robotic manipulator verify the efficiency of the proposed observer.