Robust cadence tracking for switched FES-cycling using a time-varying estimate of the electromechanical delay

To combat the severity of neurological conditions (NCs), limit the complications, and reduce the cost of treatment, researchers have turned to hybrid exoskeletons such as functional electrical stimulation (FES) cycling. In this work, closed-loop FES/motor controllers are developed that compensate for time-varying, nonlinear, and uncertain dynamics, unknown disturbances, switching between actuators (e.g., between muscle groups and the motor), fatigue, and the unknown time-varying muscle delay between stimulation application and the production of muscle force, called the electromechanical delay (EMD). Control authority is maintained and efficient muscle contractions are produced through the development of FES/motor switching conditions that are both EMD and state dependent. Contributions are that the controllers implement a modular time-varying estimate of the EMD and yield exponential cadence tracking as verified by a Lyapunov-like stability analysis. An example EMD estimate is presented that varies with cycling time to account for fatigue. Furthermore, experiments were conducted to validate the developed control system, which produced an average cadence error of -0.01 ± 1.35 revolutions per minute (RPM) across five able-bodied participants and -0.05 ± 1.38 RPM across four participants with NCs.