Linearization of an opensim arm model for feedback control design
Nenhuma Miniatura disponível
Citações na Scopus
0
Tipo de produção
Artigo de evento
Data
2019
Autores
PINHEIRO, W. C.
Castro, M.C.F.
MENEGALDO, L. L.
Castro, M.C.F.
MENEGALDO, L. L.
Orientador
Periódico
IFMBE Proceedings
Título da Revista
ISSN da Revista
Título de Volume
Citação
PINHEIRO, W. C.; CASTRO, M. C. F.; MENEGALDO, L. L. Linearization of an opensim arm model for feedback control design. IFMBE Proceedings, v. 70, n. 1, p. 289-294, 2019.
Texto completo (DOI)
Palavras-chave
Resumo
© 2019, Springer Nature Singapore Pte Ltd.One of the challenges of Biomechanics science is simulating the physiological or artificial feedback control of human joints. OpenSim is an opensource software package for simulation of musculoskeletal systems and movement analysis that provides highly detailed models suitable for forward dynamics simulations. However, the lack of access to model’s Ordinary Differential Equations explicitly hampers the design of feedback controllers based on local linear approximations of system dynamics. System identification techniques can be applied to obtain such linear model approximations and allow model-based control design. This paper aims to apply Recursive Instrumental-Variable Method (IV) as an identification technique to approximate muscle activation and contraction/joint dynamics of the biceps brachialis in an OpenSim arm model. As a result, activation dynamics and contraction/joint dynamics were identified and modeled by transfer function models with 71.91% and 97.35% of accuracy, respectively, to the nonlinear OpenSim model. It was also possible to observe how nonlinear effects should be taken into account when designing the controllers.