Velocity-Aided IMU-Based Attitude Estimation
Résumé
This paper addresses the problem of estimating the attitude of a rigid body, which is subject to high accelerations and equipped with inertial measurement unit (IMU) and sensors providing the body velocity (ex-pressed in the reference frame attached to the body). That issue can be treated differently depending on the level of confidence in the measurements of the magnetometer of the IMU, particularly with regard to the observation of the inclination component with respect to the vertical direction, rendering possible to describe the interaction with gravity. Two cases are then studied: either (i) the magnetometer is absent and only the inclination can be estimated, (ii) the magne-tometer is present, giving redundancy and full attitude observability. In the latter case, the presented observer allows to tune how much the inclination estimation is influenced by the magnetometer. All state estimators are proposed with proof of almost global as-ymptotic stability and local exponential convergence. Finally, these estimators are compared with state-of-the-art solutions in clean and noisy simulations, allowing recommended solutions to be drawn for each case.
Domaines
Automatique / Robotique
Fichier principal
benallegue2019tac_HAL_feb_2020.pdf (3.31 Mo)
Télécharger le fichier
m_l_noise.pdf (476.9 Ko)
Télécharger le fichier
plots_realstate.pdf (63.76 Ko)
Télécharger le fichier
tiltcomparison.pdf (409.59 Ko)
Télécharger le fichier
tiltcomparison_noise.pdf (457.78 Ko)
Télécharger le fichier
x2noise.pdf (531.17 Ko)
Télécharger le fichier
x_2.pdf (585.62 Ko)
Télécharger le fichier
m_l.pdf (422.92 Ko)
Télécharger le fichier
Origine | Fichiers produits par l'(les) auteur(s) |
---|
Format | Figure, Image |
---|
Format | Figure, Image |
---|
Format | Figure, Image |
---|
Format | Figure, Image |
---|
Format | Figure, Image |
---|
Format | Figure, Image |
---|