TY - GEN
T1 - Adaptive Feedforward Control of Cydic Movements Using Artificial Neural Networks
AU - Abbas, James J.
AU - Chizeck, Howard J.
N1 - Publisher Copyright:
© 1992 Institute of Electrical and Electronics Engineers Inc.. All rights reserved.
PY - 1992
Y1 - 1992
N2 - An adaptive neural network control system has been designed for the purpose of controlling cyclic movements of nonlinear dynamic systems with input time delays. The work is part of a larger project directed at the development of Functional Neuromuscular Stimulation (FNS) systems for restoring the ability to stand and to walk to people with paralysis of lower extremity musculature. The adaptive feedforward (FF) controller is implemented as a two-stage neural network. The first stage, the pattern generator (PG), generates a cyclic pattern of activity. The signals from the PG are additively filtered by the second stage, the pattern shaper (PS). This stage uses modifications to standard artificial neural network learning algorithms to adapt its filter properties. The control system has been evaluated in computer simulation on a musculoskeletal model which consisted of two muscles acting on a swinging pendulum. The control system was demonstrated to provide automated customization of the FF controller parameters for a given musculoskeletal system as well as on-line adaptation of the FF controller parameters to account for changes in the musculoskeletal system. This addictive feedforward control strategy may be impropriate for other applications in the control of nonlinear dynamic systems with input time delays.
AB - An adaptive neural network control system has been designed for the purpose of controlling cyclic movements of nonlinear dynamic systems with input time delays. The work is part of a larger project directed at the development of Functional Neuromuscular Stimulation (FNS) systems for restoring the ability to stand and to walk to people with paralysis of lower extremity musculature. The adaptive feedforward (FF) controller is implemented as a two-stage neural network. The first stage, the pattern generator (PG), generates a cyclic pattern of activity. The signals from the PG are additively filtered by the second stage, the pattern shaper (PS). This stage uses modifications to standard artificial neural network learning algorithms to adapt its filter properties. The control system has been evaluated in computer simulation on a musculoskeletal model which consisted of two muscles acting on a swinging pendulum. The control system was demonstrated to provide automated customization of the FF controller parameters for a given musculoskeletal system as well as on-line adaptation of the FF controller parameters to account for changes in the musculoskeletal system. This addictive feedforward control strategy may be impropriate for other applications in the control of nonlinear dynamic systems with input time delays.
UR - http://www.scopus.com/inward/record.url?scp=85012282361&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85012282361&partnerID=8YFLogxK
U2 - 10.1109/IJCNN.1992.226884
DO - 10.1109/IJCNN.1992.226884
M3 - Conference contribution
AN - SCOPUS:85012282361
T3 - Proceedings of the International Joint Conference on Neural Networks
SP - 832
EP - 837
BT - Proceedings - 1992 International Joint Conference on Neural Networks, IJCNN 1992
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 1992 International Joint Conference on Neural Networks, IJCNN 1992
Y2 - 7 June 1992 through 11 June 1992
ER -