Reflex Response Modelling of Exoskeleton-User Interaction

Abstract

An exoskeleton can help older aged adults stay independent for a longer time by augmenting their torque at weakened joints. Predictive models are useful to develop an exoskeleton and assess the effect of exoskeleton s on the user. Existing models that predict the effect of the exoskeleton on the user use a pre-defined motion which does not change when the forces applied by the exoskeleton change. In our work, we capture the effect of the exoskeleton on the user’s motion by a fitted muscle reflex model and a musculoskeletal model. The model explains the motion deviation with respect to a base motion, as opposed to a steady pose. In an experiment a base motion is perturbed with an external torque. The resulting predicted motion correlates well with the measured motion. As the response of a mechanical system due to forces is well known, the dynamic interaction, incorporating changes in the motion, between an exoskeleton and its user can be predicted with the reflex model.