An adaptive arm’s mechanical impedance estimator for rehabilitation robots without force and acceleration sensors

Abstract

The mechanical impedance of the human arm can be measured to quantitatively assess the motor function. In this paper, an adaptive least square estimation method is proposed for online measurement of the mechanical impedance of the arm, without force or acceleration sensors, which extremely reduces the expenses and complexity of rehabilitation robots. The proposed strategy may also be used for monitoring the dynamic changes of the mechanical impedance. Estimation of time-varying force is also the other capability of the algorithm. The closed-loop stability of the system is analytically shown using the Lyapunov stability theorem. To show the applications of the proposed scheme, two main scenarios are described which can be used for the rehabilitation robots to assess the motor recovery of the patients undergoing the rehabilitation sessions. To validate the scheme, a 2-DOF manipulator is adopted to illustrate the accuracy of the impedance and force estimations in noiseless and noisy conditions.