COMPARISON OF SIMULATION AND EXPERIMENTS OF DIFFERENT CLOSED-LOOP STRATEGIES FOR FUNCTIONAL ELECTRICAL-STIMULATION - EXPERIMENTS IN PARAPLEGICS

Citation
J. Quintern et al., COMPARISON OF SIMULATION AND EXPERIMENTS OF DIFFERENT CLOSED-LOOP STRATEGIES FOR FUNCTIONAL ELECTRICAL-STIMULATION - EXPERIMENTS IN PARAPLEGICS, Artificial organs, 21(3), 1997, pp. 232-235
Citations number
7
Categorie Soggetti
Engineering, Biomedical
Journal title
ISSN journal
0160564X
Volume
21
Issue
3
Year of publication
1997
Pages
232 - 235
Database
ISI
SICI code
0160-564X(1997)21:3<232:COSAEO>2.0.ZU;2-E
Abstract
Open-loop and closed-loop stimulation of the knee extensors for the co ntrol of the knee joint angle and torque were tested as a potential ba sis for more complex functional electrical stimulation (FES) systems t o be used in human locomotion. The output of the biomechanical simulat ion model described previously was compared with stimulation experimen ts in patients with complete thoracic spinal cord injury. Good corresp ondence between simulation and experiments was obtained under both iso metric conditions and conditions with a freely swinging shank. For clo sed-loop control, a simple proportional integral derivative (PID) cont roller yielded sufficient performance only under isometric conditions, especially if combined with (linear) feedforward. Because of addition al nonlinearities of musculotendon and body segmental dynamics, more c omplex strategies have to be applied to the control of unconstrained m ovements. To compensate for these nonlinearities, an inverse model was derived from the direct biomechanical model. This inverse model had s atisfactory agreement between the measured knee angle and the desired trajectory already under open-loop conditions. A combination of the in verse model in the feedforward part of the control loop and a PID cont roller provided robust and precise control of the knee angle. Further improvement may be achieved by including elements of spasticity into t he simulation model and by controlling both agonistic and antagonistic muscles.