Position and cadence tracking of a motorized FES-cycle with an unknown time-varying input delay using saturated FES control

Abstract

Neurological conditions (NCs) can have debilitating effects on an individual and can impede a person’s ability to perform activities of daily living. A common rehabilitation for individuals with NCs is the use of functional electrical stimulation (FES) to induce muscle contractions to perform a functional task, such as cycling. In this paper, we develop a closed-loop FES cycling controller that yields exponential cadence tracking. Contributions of the work result from the design and analysis innovations to compensate for switching between the different muscles and motor control inputs, compensating for the inherent uncertain nonlinear dynamics, compensating for the time-varying unknown input delay resulting from the complex electrochemical FES muscle torque production process, and compensating for saturation in the evoked torque due to fatigue, available muscle mass, or stimulation sensitivity. The performance of the developed control system was examined through a series of experiments on nine participants (five able-bodied and four with NCs). The experiments resulted in a cadence tracking error of −0.03 ± 1.69 RPM for non-informed passive able-bodied participants and −0.04 ± 1.98 RPM for the participants with NCs, for a desired cadence of 50 RPM.