Distributed multi-source information fusion algorithm for vehicle tracking with priori uncertainty

Zhenliang Ma, Jianping Xing, Junchen Sha, Liang Gao, Songpu Ai

Research output: Contribution to journalArticleResearchpeer-review

Abstract

Maximizing multi-source information content has already been recognized as an effective way to achieve robust and accurate tracking performance in the vehicle tracking field. The complex tracking environment and uncertain vehicle states bring much uncertainty to the tracking system, which make it is difficult to track vehicle accurately. In this paper, a distributed multi-source information fusion algorithm for vehicle tracking with priori uncertainty is proposed. The norm-bounded uncertainty exists in both system and observation matrix, and there is no assumption of the noise characteristic. According to the H control theory and the convex optimization method, a distributed fusion algorithm is presented by solving LMIs (Linear Matrix Inequalities). Finally, vehicle tracking experiments are provided to demonstrate the accuracy and robustness of the proposed tracking algorithm by comparing with the commonly used Kalman Filter (KF) and Particle Filter (PF) method. The influence of the system uncertainties to the proposed tracking algorithm performance is also investigated.

Original languageEnglish
Pages (from-to)853-862
Number of pages10
JournalJournal of Information and Computational Science
Volume9
Issue number4
Publication statusPublished - 1 Apr 2012

Keywords

  • Distributed multi-source information fusion
  • H
  • LMI
  • Priori uncertainty
  • Vehicle tracking

Cite this