A motion assistance strategy for upper limb robotic rehabilitation based on dual-person competitive behavioral performance

This article addresses the issue of imbalanced task difficulty and motion intensity caused by differences in patients’ motor abilities during dual-person upper-limb competitive interactive training. An adaptive motion assistance strategy based on behavioral performance is proposed. Two kinematic coefficients are introduced to quantify behavior performance, and an adaptive adjustment algorithm is used to dynamically regulate assistance levels and virtual spring stiffness coefficients. To validate the proposed strategy, multiple experiments were conducted under the competitive interaction mode designed based on force haptic feedback. The results demonstrate that the strategy can adaptively adjust the level of assistance based on behavior performance, effectively balancing individual task difficulty and motion intensity differences. This provides a dynamic and efficient motion assistance method for dual-person competitive interactive training, further exploring the potential of social interaction in robot-assisted rehabilitation.