Skip to Main content Skip to Navigation
Theses

Contribution à l'identification des paramètres inertiels des segments du corps humain

Abstract : In this work, an identification method is presented for the determination of the body segment parameters. The identification consists in writing the reaction model: the 6 components of the interaction between the multibody system and outside are put in relation with the segments kinematic. The equations of motion are written in matrix form. The 6 components of the external forces and torques are gathered in a 4$\times$4 antisymmetric matrix and the inertial parameters are gathered in a symmetric and positive definite matrix. In order to estimate the coefficients of this matrix, a projected conjugate gradient algorithm is developed.\\
The method of identification is first applied to a rigid body. The results obtained from simulated movements allow validating the method and also presenting a simple experimental design for identifying the inertial characteristics of a real body.\\
An automatic generation of the equations of motion is then proposed for a multibody system composed of $n$ segments. The validation of the identification method in the case of a multibody system is obtained from simulated movements. Results from noisy kinematic highlight the importance of the position of the joint centers to obtain a precision of 4\%. The joints centers are then determined experimentally for 8 subjects with a functional method requiring the calculation of rotation matrices. These are obtained by an algorithm of minimization. Finally, the inertial parameters of the right upper limb of 8 subjects are identified and compared to a conventional model.
Complete list of metadatas

Cited literature [130 references]  Display  Hide  Download

https://tel.archives-ouvertes.fr/tel-00202452
Contributor : Tony Monnet <>
Submitted on : Monday, January 7, 2008 - 8:56:27 AM
Last modification on : Wednesday, March 20, 2019 - 2:14:09 PM
Long-term archiving on: : Tuesday, April 13, 2010 - 4:22:39 PM

Identifiers

  • HAL Id : tel-00202452, version 1

Collections

Citation

Tony Monnet. Contribution à l'identification des paramètres inertiels des segments du corps humain. Mécanique [physics.med-ph]. Université de Poitiers, 2007. Français. ⟨tel-00202452⟩

Share

Metrics

Record views

586

Files downloads

1864