Virtual decomposition control for flexible-joint robot manipulators with extended Kalman filtering

Hua-shan LIU,Sheng LI,Bao-xiang WANG
DOI: https://doi.org/10.7641/CTA.2017.70367
2018-01-01
Abstract:A general control approach based on virtual decomposition control (VDC) and extended Kalman filtering (EKF) is proposed in this paper, to realize trajectory tracking control for flexible-joint robot manipulators via non-full-state feedback. First, considering parametric uncertainties in robot model and external disturbances, virtual-decomposed control law for rigid-link subsystems and flexible-joint subsystems are designed, respectively. Then, to break through the limitations that the existing VDC method always needs full-state feedback, an EKF based indirect state observer is presented, to make the closed-loop control with only position and velocity measurements from the motor side of each joint, and without any state measurements from the link side. In addition, strict stability analysis of the system is given according to the theory of virtual stability and Lyapunov stability. Finally, simulation example verifies the effectiveness of the proposed approach, and comparison results show its superior tracking performance than the other typical controller based on the traditional Lagrange integrated dynamic model.
What problem does this paper attempt to address?