Adaptive Control of Flexible Joint Manipulators with Parameter Convergence

This paper is focused on the adaptive control design for flexible joint robots with parameter uncertainties. Based on the singular perturbation theory, the flexible joint robot is transformed into a fast subsystem and a slow subsystem. For the fast subsystem, the PD control strategy is used to suppress elastic oscillations by the feedback of the motor angular displacement error. For the slow subsystem, an adaptive control scheme is proposed to guarantee the parameter estimation error convergence. The closed-loop system is testified to be globally exponentially stable via the designed Lyapunov function. The comprehensive simulation results demonstrate great tracking precision and high robustness of the proposed control scheme.