Dynamical Design of ADRC for One-DOF Manipulator Systems with Time-varying Disturbance and Large Parametric Uncertainty

Abstract

In this paper, the control problem of one-degree-of-freedom robot manipulato In this paper, the control problem of one-degree-of-freedom robot manipulator systems with time-varying disturbance and significant parametric uncertainty is investigated. For the traditional linear active disturbance rejection control (ADRC), the choice of the nominal values of system parameters can signally influence the closed-loop stability. To achieve stronger robustness to the parametric uncertainties in manipulator systems, a new ADRC with dynamical design of control input is proposed. The proposed design contains an extended state observer to estimate total disturbance, and a dynamical system of control input to achieve the ideal input. Since the proposed ADRC only requires the signs of system parameters rather than the nominal values, a stronger capability of handling parametric uncertainties is shown. Finally, the comparative simulations of traditional ADRC and new design for manipulator system with three typical external disturbances and large parametric uncertainties are presented. When the closed-loop performance of the traditional ADRC becomes poor due to large parametric uncertainties, the proposed ADRC still results in a satisfied closed-loop performance. r systems with time-varying disturbance and significant parametric uncertainty is investigated. For the traditional linear active disturbance rejection control (ADRC), the choice of the nominal values of system parameters can signally influence the closed-loop stability. To achieve stronger robustness to the parametric uncertainties in manipulator systems, a new ADRC with dynamical design of control input is proposed. The proposed design contains an extended state observer to estimate total disturbance, and a dynamical system of control input to achieve the ideal input. Since the proposed ADRC only requires the signs of system parameters rather than the nominal values, a stronger capability of handling parametric uncertainties is shown. Finally, the comparative simulations of traditional ADRC and new design for manipulator system with three typical external disturbances and large parametric uncertainties are presented. When the closed-loop performance of the traditional ADRC becomes poor due to large parametric uncertainties, the proposed ADRC still results in a satisfied closed-loop performance.