Gps-free localization based on multiple imus and federated filter fusion

Jiyeon Kim, Moogeun Song, Jaehoon Kim, Dongik Lee

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

Localization techniques based on the integration of MEMSs(Micro Electro Mechanical Systems) and GPS(Global Positioning System) have been widely used in various fields, including aerospace, biomedical, automotive, and robotics, among others. However, it is well known that GPS signals are weakened or blocked in tunnels, urban canyons, or indoor environments. A solution is to use multiple IMU(Inertial Measurement Unit) sensors. In this paper, we present a federated fusion algorithm for GPS-free localization using multiple MEMS IMUs. The federated filter uses a master fusion algorithm to combine the outputs of independent sensors so that the current position can be estimated. The proposed structure exploits a quaternion-based extended Kalman filter and a linear Kalman filter for attitude estimations and velocity estimations, respectively. The effectiveness of the proposed algorithm was demonstrated with MATLAB simulation.

Original languageEnglish
Pages (from-to)708-714
Number of pages7
JournalJournal of Institute of Control, Robotics and Systems
Volume26
Issue number9
DOIs
StatePublished - 2020

Keywords

  • Extended Kalman filter
  • Federated filter
  • Multiple IMUs
  • Position estimation
  • Quaternion

Fingerprint

Dive into the research topics of 'Gps-free localization based on multiple imus and federated filter fusion'. Together they form a unique fingerprint.

Cite this