Integrated SVD/EKF for small satellite attitude determination and rate gyro bias estimation

Research output: Contribution to journalConference articlepeer-review

10 Citations (Scopus)


In this study an integrated singular value decomposition (SVD)/ Extended Kalman filter (EKF) attitude determination system is presented, in which the SVD and EKF algorithms are combined to estimate the attitude angles and angular velocities and gyro biases, respectively. As a reference directions for SVD, the unit vectors toward the Sun and Earth's Magnetic Field are used. The Euler angles produced SVD are provided as input to the EKF along with angular rate data. The parameters of satellite's rotational motion and rate gyro biases are estimated using EKF. In comparison to more traditional approaches, this preprocessing step significantly reduces the complexity of filter design by allowing the use of linear measurement equations.

Original languageEnglish
Pages (from-to)233-238
Number of pages6
Issue number9
Publication statusPublished - 1 Jul 2015
Event1st IFAC Workshop on Advanced Control and Navigation for Autonomous Aerospace Vehicles, ACNAAV 2015 - Seville, Spain
Duration: 10 Jun 201512 Jun 2015

Bibliographical note

Publisher Copyright:
© 2015, IFAC (International Federation of Automatic Control) Hosting by Elsevier Ltd. All rights reserved.


The work was jointly supported by TUBITAK (The Scientific and Technological Research Council of Turkey), Grant 113E595, and RFBR (Russian Foundation for Basic Research), Grant 14-08-91373.

FundersFunder number
Russian Foundation for Basic Research14-08-91373
Türkiye Bilimsel ve Teknolojik Araştirma Kurumu113E595


    • Angular velocity
    • Attitude determination
    • Estimation
    • Extended kalman filter
    • Magnetometer
    • Rate gyro
    • Singular value decomposition
    • Small satellite
    • Sun sensor


    Dive into the research topics of 'Integrated SVD/EKF for small satellite attitude determination and rate gyro bias estimation'. Together they form a unique fingerprint.

    Cite this