Distributed fault-tolerant fusion estimation based on multiple-model extended kalman filter

Xiaodi Shi, Liping Yan*, Yuanqing Xia

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

As practical nonlinear systems become progressively complex, the extended Kalman filter (EKF) is limited for many applications because of its performance degradation. In this paper, we propose a novel multiple-model extended Kalman filter (MMEKF), which remarkably reduced its large deviation. The expansion points designed in the MMEKF algorithm obey the Gaussian distribution in the process of probabilistic models, which are used to approximately represent the whole state space by using multiple probabilistic weighted method. Compared with other filters such as EKF, UKF and CKF, the MMEKF shows higher estimation accuracy for unreliable measurements especially in multi-sensor systems. This paper also considers fault-tolerant distributed data fusion estimation, whose feasibility and effectiveness through a numerical example.

Original languageEnglish
Title of host publicationProceedings of the 38th Chinese Control Conference, CCC 2019
EditorsMinyue Fu, Jian Sun
PublisherIEEE Computer Society
Pages3450-3455
Number of pages6
ISBN (Electronic)9789881563972
DOIs
Publication statusPublished - Jul 2019
Event38th Chinese Control Conference, CCC 2019 - Guangzhou, China
Duration: 27 Jul 201930 Jul 2019

Publication series

NameChinese Control Conference, CCC
Volume2019-July
ISSN (Print)1934-1768
ISSN (Electronic)2161-2927

Conference

Conference38th Chinese Control Conference, CCC 2019
Country/TerritoryChina
CityGuangzhou
Period27/07/1930/07/19

Keywords

  • Distributed fusion
  • Multiple-model extended Kalman filter
  • Probabilistic model design
  • Unreliable measurements

Fingerprint

Dive into the research topics of 'Distributed fault-tolerant fusion estimation based on multiple-model extended kalman filter'. Together they form a unique fingerprint.

Cite this