Cycling exercise - based on functional electrical stimulation (FES) for disabled individuals - is achieved by stimulating single muscle group, the quadriceps, with the help of a new assist mechanism represented by a solid disc flywheel equipped with an electrical clutch. Fuzzy logic based closed-loop control method is implemented to obtain a stable cycling cadence by i) controlling the stimulation intensity on the muscle and ii) managing the engagement of the flywheel mechanism. To achieve better results, several crank positions - with different gear ratios between the crank and the flywheel - are tested and analyzed. To obtain the optimal design parameters, a multi-objective genetic algorithm (MOGA) approach is adopted towards maximizing the cycling efficiency and minimizing the cadence error. A comparison with the results reported in the literature reveals the superiority of the proposed design to limit the cadence error to ±5 rpm for 35 rpm desired speed. Moreover, the results demonstrate that the designed control approach with the proposed assist mechanism is robust to changes in muscle force due to muscle fatigue. Additionally, the introduced control approach with the new assist mechanism is promoting bounded tracking of the desired speed and prolonging FES-cycling training by stimulating the quadriceps muscle group only.
Key words: Cycling aid mechanism; Functional electrical stimulation; Cadence control; Multi-objective genetic algorithm; Fuzzy control.
|