摘要
This study presents the development of a motion-decoupled finger exoskeleton designed both to facilitate passive rehabilitation and enable joint stiffness estimation. The proposed exoskeleton incorporates a self-alignment mechanism to accommodate variations in joint alignment and finger dimensions, ensuring an optimal fit for diverse users. Furthermore, linkage optimization was implemented to enhance force transmission efficiency, thereby improving transmission angles and reducing mechanical resistance. To achieve accurate joint stiffness estimation, a force and friction model was developed to compensate for frictional effects. In this study, a series of experimental validations was conducted using prosthetic finger models equipped with six torsional springs of varying stiffness in order to determine friction model parameters and evaluate measurement accuracy. The results indicate that the exoskeleton can simultaneously estimate joint stiffness during constant-velocity passive motion, achieving an approximate error of 5.55 % for stiffness values corresponding to nonzero levels on the Modified Ashworth Scale. These findings suggest that the proposed exoskeleton provides a reliable platform for joint stiffness assessment in rehabilitation settings and holds significant potential for clinical applications in stroke recovery as well as musculoskeletal rehabilitation.