Skip to Main content Skip to Navigation
Theses

Estimation des caractéristiques du mouvement humain en station debout. Mise en œuvre d'observateurs flous sous forme descripteur

Abstract : Neurophysiology and biomechanics brings two complementary approaches to the postural control scheme. The human body, which is controlled by the central nervous system, is considered as rigid body segments connected by hinge joints system where inputs are the joint torques and the outputs are the segment positions. Due to the postural strategies employed in standing, a double inverted pendulum nonlinear model is chosen. It depends on parameters that can be estimated by standard anthropometric tables. In order to adjust the model for an individual, a simulated annealing with heating cycles is chosen. Joint torques are classically estimated by the well known uncertain inverse dynamics techniques. An alternative approach, stemming from control theory, is then proposed. That one consists on the use of unknown inputs observers. A class of fuzzy observers based on a descriptor form is then studied. The convergence conditions of the prediction error are obtained via a Lyapunov quadratic function. These ones can be conservative. Four LMI or BMI based approaches are then proposed to relax these conditions. In the case of human stance, the results obtained with the use of an unknown input observer are compared with the ones obtained with the use of different inverse dynamics approaches. It seems that observer based results are more accurate for human standing studies. Moreover, the real time character of the observer based approach is underlined and leads to many rehabilitation applications.
Document type :
Theses
Complete list of metadatas

Cited literature [19 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00007960
Contributor : Kevin Guelton <>
Submitted on : Friday, January 7, 2005 - 4:29:17 PM
Last modification on : Friday, November 13, 2020 - 8:44:11 AM
Long-term archiving on: : Friday, April 2, 2010 - 9:26:01 PM

Identifiers

  • HAL Id : tel-00007960, version 1

Citation

Kevin Guelton. Estimation des caractéristiques du mouvement humain en station debout. Mise en œuvre d'observateurs flous sous forme descripteur. Automatique / Robotique. Université de Valenciennes et du Hainaut-Cambresis, 2003. Français. ⟨tel-00007960⟩

Share

Metrics

Record views

646

Files downloads

2004