Fuzzy and learning control of FES induced gait

Abstract Research is done to restore gait by means of functional electrical stimulation For people suffering from paraplegia. Difficulties are that muscle fatigue and heavily changing patient conditions deteriorate the performance of such systems. The design of feedback controllers is considered that adapt stimulation parameters to compensate for this deterioration. Conventional control (PID) gives good performance when muscle and leg dynamics are known, but requires recurring identification. By using fuzzy controllers based on general, qualitative knowledge, and by providing self-tuning capabilities through learning, identification might be avoided. Simulation studies confirm this.