Abstract

The creation of a human dynamical model useful in upper-limb exoskeleton control remains an open problem. We present a framework that approaches model generation from a “sensor-driven” design perspective that explicitly avoids over- fitting parameters and minimally relies on literature values and biological assumptions. Initial results on synthetic data for a simplified model of the elbow indicate that this framework is a viable starting point from which to build more sophisticated dynamical models. Full results can be found in Tech Rep. UCB/EECS-2016-66.