Event-Triggered Reduced-Order Filtering for Continuous Semi-Markov Jump Systems with Imperfect Measurements

Huiyan Zhang, Hao Sun, Xuan Qiu, Rongni Yang, Shuoyu Wang, Ramesh K. Agarwal

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

This article conducts the issue of event-triggered reduced-order filtering for continuous-time semi-Markov jump systems with imperfect measurements as well as randomly occurring uncertainties (ROUs). Specifically, the sojourn-time-dependent transition probability matrix (TPM) is presumed to be polytopic and a quantizer is introduced to quantize output signals aiming to reflect the reality. Both ROUs and sensor failures are generated by individual random variables belonging to be mutually independent Bernoulli-distributed white sequences. First, sufficient conditions for the existence of the event-triggered reduced-order filter are obtained by utilizing the dissipativity-based technique to ensure the asymptotical stability with a strictly dissipative performance of the filtering error system. The time-varying TPM is then fractionalized, which enhances the results as stated. Furthermore, the required reduced-order filter parameters are obtained by introducing slack symmetric matrix as well as cone complementarity linearization algorithm. The effectiveness of the suggested event-triggered reduced-order filter design method is shown through simulation results.

Original languageEnglish
Pages (from-to)6145-6157
Number of pages13
JournalIEEE Transactions on Cybernetics
Volume54
Issue number10
DOIs
StatePublished - 2024

Keywords

  • Event-triggered control
  • quantization
  • reduced-order filtering
  • semi-Markov jump systems (SMJSs)
  • sensor failures

Fingerprint

Dive into the research topics of 'Event-Triggered Reduced-Order Filtering for Continuous Semi-Markov Jump Systems with Imperfect Measurements'. Together they form a unique fingerprint.

Cite this