Disturbance rejection tracking control of robotic manipulators based on extended state observer

Abstract

The trajectory tracking control problem of multi-degree-of-freedom robotic manipulators ((DOF-RM) has attracted widespread attention due to inherent challenges, including nonlinearity, parameter uncertainty, and unknown external disturbances. This paper proposes a robust trajectory tracking control scheme utilizing linear extended state observer (LESO). First, a dynamic model of the robotic manipulator is established, and all unmodeled dynamics and exogenous perturbations are merged into a lumped disturbance term. Second, a parameterized high-gain LESO is designed for realtime estimation of the system state and lumped disturbances, the designed observer only requires the adjustment of a single bandwidth parameter. Finally, the LESO is combined with a robust tracking controller to design an active disturbance-resistant tracking controller. Rigorous stability proofs using Lyapunov stability theory demonstrate that the tracking error and the observer estimation error are uniformly and eventually bounded. Finally, numerical simulations of a two-DOF robotic arm verify the tracking performance of the proposed method.