Contacting mechanical impedance of human finger based on uncertain system

For hand rehabilitation after stroke, exoskeletons are applied to replace the traditional therapy which relies on experience too much. However, as sensors couldn´t be placed between finger and object, we have to estimate the haptic force by the impedance model. Conventional methods describing the impedance as a spring-damper-mass system can´t fit the experiment precisely. In this paper, we propose a method to describe the contacting impedance based on uncertain system. The physical model is a bilateral contacting model imitating the finger anatomy. The varying impedance of finger is considered as perturbation. We set experiments under three main postures of human finger. By comparison, the bionic model has been proved more efficient. The experiments give out the nominal value and range of the parameters, according to which we construct the perturbation weighting function. Finally, the uncertain system of contacting impedance of human finger is presented, that shows the mechanical impedance more fully. So we can apply the robust control theory to control exoskeleton conveniently.