DocumentCode
3126486
Title
Adaptive Knee Joint Control for an Active Amputee Prosthesis
Author
Scandaroli, Glauco Garcia ; Borges, Geovany Araújo ; Da Rocha, Adson F. ; Nascimento, Francisco Assis de Oliveira
Author_Institution
Dept. of Electr. Eng., Univ. de Brasilia, Brasilia
fYear
2008
fDate
29-30 Oct. 2008
Firstpage
164
Lastpage
169
Abstract
This work presents the design and evaluation of knee joint controllers for an above-knee prosthesis. The prototype is a three degree of freedom prosthesis which is planned to be controlled in part by EMG thigh signals, in part by a predictive trajectory generator based on multisensorial data such that the foot can be correctly positioned with respect the floor. There is described a simplified non-linear dynamic model for the knee opening angle. The model parameter identification is made using a least squares batch estimator. Based on the estimated non-linear model, two strategies for joint controllers are designed: PID and model reference adaptive controller. The PID design follows a root locus approach with experimental tuning. Preliminary tests were performed in a platform built for the prosthesis under development. The results discussed in this paper reveals the difficulty in controlling such a nonlinear plant.
Keywords
adaptive control; control system synthesis; electromyography; least squares approximations; medical robotics; medical signal processing; nonlinear dynamical systems; parameter estimation; patient rehabilitation; position control; prosthetics; three-term control; EMG thigh signal processing; PID design; active amputee prosthesis; adaptive knee joint controller design; least square batch estimator; multisensorial data; nonlinear dynamic model; parameter estimation; predictive trajectory generator; rehabilitation robotics; Adaptive control; Electromyography; Foot; Knee; Programmable control; Prosthetics; Prototypes; Signal generators; Thigh; Trajectory; above-knee prosthesis; adaptive control; rehabilitation robotics.;
fLanguage
English
Publisher
ieee
Conference_Titel
Robotic Symposium, 2008. LARS '08. IEEE Latin American
Conference_Location
Natal, Rio Grande do Norte
Print_ISBN
978-1-4244-3379-7
Electronic_ISBN
978-0-7695-3536-4
Type
conf
DOI
10.1109/LARS.2008.12
Filename
4812643
Link To Document