Adaptive Robust Control With Extended State Observer For Human-Robot Impedance

Human-robot systems have tremendous potentials in engineering applications. To be ubiquitous, the robot should show particular impedance according to the planned action. A novel adaptive robust controller (ARC) with extended state observer (ESO) based on output feedback in the Cartesian coordinates is proposed aiming at guaranteeing the target human-robot interaction impedance, which needs to be regulated at hand. The adaptive robust control is designed to cope with the parametric uncertainties and unknown nonlinearities. An extended state observer is incorporated to address the challenging issue that the human-robot system is subjected to various external disturbances. The integration of the two algorithms is derived to weaken the control chattering and improve the tracking performances. Tracking differentiator (TD) is performed to achieve output feedback when some states are unavailable. The proof of the control framework stability is provided by the Lyapunov stability theorem. Furthermore, the effectiveness of the proposed controller is demonstrated by simulations and experiments. The results show that the adaptive robust control with extended state observer (ARC with ESO) offers superior convergence and satisfactory tracking performances. The tracking error is decreased and the control chattering is weakened due to the disturbance estimation and compensation by the extended state observer. The proposed controller has significant application potentials in the fields of man- machine interaction.