Energy-aware Adaptive Attitude Estimation Under External Acceleration for Pedestrian Navigation - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Journal Articles IEEE/ASME Transactions on Mechatronics Year : 2016

Energy-aware Adaptive Attitude Estimation Under External Acceleration for Pedestrian Navigation

Hassen Fourati
Alain Kibangou

Abstract

In this paper, we consider the problem of rigid bodyattitudeestimationunderexternalaccelerationusingasmallinertial/magneticsensors module containing a triad of gyroscope, accelerometer,andmagnetometer.Thepaperisfocusedontwomainchallenges. The first challenge concerns the attitude estimationduring dynamic case, in which external acceleration occurs. Thislatter leads to lose performance in attitude estimation methods. Aquaternion-based adaptive Kalman filter (q-AKF) compensatingexternal acceleration from the residual in the accelerometer isdesigned. At each step, the covariance matrix of the externalacceleration is estimated to tune the filter gain adaptively. Thesecond challenge is related to the energy consumption issue ofgyroscope. In order to ensure a longer battery life for the InertialMeasurement Units (IMUs), we study the way to reduce the gyromeasurements acquisition by switching on/off the sensor whilemaintaining an acceptable attitude estimation. A smart detectionapproach isproposed to decide whether the body is indynamic orstatic case. The efficiency of the q-AKF is demonstrated throughnumerical simulations and experimental tests.
Fichier principal
Vignette du fichier
MFK_final_version.pdf (2.08 Mo) Télécharger le fichier
Origin : Files produced by the author(s)
Loading...

Dates and versions

hal-01241403 , version 1 (24-11-2016)

Identifiers

Cite

Aida Makni, Hassen Fourati, Alain Kibangou. Energy-aware Adaptive Attitude Estimation Under External Acceleration for Pedestrian Navigation. IEEE/ASME Transactions on Mechatronics, 2016, 21 (3), pp.1366-1375. ⟨10.1109/TMECH.2015.2509783⟩. ⟨hal-01241403⟩
423 View
358 Download

Altmetric

Share

Gmail Facebook X LinkedIn More