Multiple model switched repetitive control for tremor suppression

Abstract

Tremor is a condition that impacts millions of people globally, and is characterised by a periodic limb movement that impedes voluntary motion. Recent studies have shown that functional electrical stimulation (FES) can help reduce tremor by artificially stimulating opposing muscles, thereby decreasing the oscillation’s amplitude. Various control methods have been proposed for this purpose, but repetitive control (RC) has shown the most promise with potential to completely suppress the tremor. While several RC approaches have demonstrated suppression rates of up to 90%, they heavily rely on an accurate model of the underlying dynamics, and their effectiveness declines steeply due to factors like muscle fatigue, spasticity, and modelling inaccuracies.

This paper introduces a multiple model switched repetitive control (MMSRC) framework that addresses the limitations of existing RC approaches. It guarantees high performance tremor suppression provided the true dynamics belong to an uncertainty set specified by the designer. This enables it to adapt to time-varying physiological changes, as well as changes in the placement of the FES electrodes. Moreover, once an uncertainty set has been established, it removes the need for subsequent model identification. This is an important step towards home-based tremor suppression where model identification and expert tuning are not possible. Experimental validation is performed with four participants, showing that MMSRC effectively suppresses tremor even in the presence of severe modelling uncertainty and fatigue, unlike conventional RC methods which often become unstable under these conditions.