Parameter Identification and Swing-up Control of an Acrobot System

Abstract

This paper considers parameter identification and swing-up balancing control of a 2-link underactuated acrobat robot called an Acrobot. For parameter identification of the Acrobot, we propose a new method using the unscented Kalman filter (UKF). The effectiveness of this method has been clarified through a parameter identification experiment. Furthermore, we studied the Acrobot's swing-up and balancing control by applying the energy based control for swing-up control and the Lyapunov based scheduling procedure for balancing control. Simulation and experimental results indicated that this approach is suitable for control of the Acrobot.