Multiple–Model UKF/CKF State Estimation for Nonlinear Systems

Xiaodi Shi, Liping Yan*, Yuanqing Xia, Bo Xiao

*Corresponding author for this work

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

Abstract

In most control systems, modeling error and noise interference will always lead to the performance degradation and divergence of the UKF or the CKF. To settle a matter caused by model uncertainties, a new UKF/CKF frame combined with multiple model method is presented in this paper. Through probabilistic multiple model design method, this paper approximates the posterior densities by a finite number of probabilistically weighted points and uses these points to display the entire state space. Simulation results and comparison analysis demonstrate that the multiple-model UKF(MMUKF) and the multiple-model CKF(MMCKF) have higher precision and stronger robustness than the traditional UKF and CKF in case of model uncertainties.

Original languageEnglish
Title of host publicationAdvances in Guidance, Navigation and Control - Proceedings of 2020 International Conference on Guidance, Navigation and Control, ICGNC 2020
EditorsLiang Yan, Haibin Duan, Xiang Yu
PublisherSpringer Science and Business Media Deutschland GmbH
Pages79-90
Number of pages12
ISBN (Print)9789811581540
DOIs
Publication statusPublished - 2022
EventInternational Conference on Guidance, Navigation and Control, ICGNC 2020 - Tianjin, China
Duration: 23 Oct 202025 Oct 2020

Publication series

NameLecture Notes in Electrical Engineering
Volume644 LNEE
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Conference

ConferenceInternational Conference on Guidance, Navigation and Control, ICGNC 2020
Country/TerritoryChina
CityTianjin
Period23/10/2025/10/20

Keywords

  • Multiple model CKF (MMCKF)
  • Multiple model UKF (MMUKF)
  • Probabilistic model design
  • Probabilistically weighted points

Fingerprint

Dive into the research topics of 'Multiple–Model UKF/CKF State Estimation for Nonlinear Systems'. Together they form a unique fingerprint.

Cite this