Human-robot interaction force information is of great significance for realizing safe, compliant and efficient rehabilitation training.
In order to accurately estimate the interaction force during human-robot
interaction, an adaptive method for estimation of human-robot interaction force is proposed in this paper. Firstly, the dynamics of humanrobot system are modeled, which allows to establish a state space equation. Then, the interaction force is described by a polynomial function
of time, and is introduced into the state space equation as a system
state. Meanwhile, the Kalman filter is adopted to estimate the extended
state of system online. Moreover, in order to deal with the uncertainty
of system noise covariance matrix, sage-husa adaptive Kalman filter is
used to correct the covariance matrices of system noises online. Finally,
experiments were carried out on a lower limb rehabilitation robot, and
the results show that the proposed method can precisely estimate the
interaction force and also has good real-time performance.