Sensitivity analysis and identification of human parameters for an adaptive, underactuated hand exoskeleton

Abstract

Designing an underactuated hand exoskeleton system that contains user’s finger in the kinematics loop improves the adjustability for different hand sizes. On the downside, the configuration and motion level kinematics become highly responsive to the finger properties and how the device is worn. In this paper, first we investigate the sensitivity of these variable parameters on the finger pose and force transmission analysis. Considering that the manual measurements of these parameters might be inaccurate and time consuming, we present an adaptive calibration algorithm to estimate them when a user is asked to wear the device. Finally, comparing the results of the calibration algorithm to the readings of an optical tracking system shows that the proposed solution is capable of estimating the variable parameters well enough to improve the kinematics analysis.