A sliding mode observer for robust fault reconstruction in a class of nonlinear non-infinitely observable descriptor systems

Joseph Chang Lun Chan, Tae H. Lee, Chee Pin Tan

Research output: Contribution to journalArticleResearchpeer-review

54 Citations (Scopus)

Abstract

This paper presents a sliding mode observer (SMO) for robustly reconstructing faults affecting a class of nonlinear non-infinitely observable descriptor systems. Preliminary transformations are utilised to re-express the system such that the design freedom in its structure is easier to exploit. An infinitely observable system is formed by treating some states as unknown inputs. The faults are then reconstructed from measurable inputs and outputs using a SMO. The structure of the SMO bounds the nonlinear components of its estimation errors. The effect of disturbances on the fault reconstruction is minimised using linear matrix inequality techniques. A summarised design procedure for the SMO scheme is presented. Finally, a simulated example is carried out to showcase the efficacy of the scheme.

Original languageEnglish
Pages (from-to)1023-1036
Number of pages14
JournalNonlinear Dynamics
Volume101
Issue number2
DOIs
Publication statusPublished - Jul 2020

Keywords

  • Descriptor systems
  • Fault reconstruction
  • Infinite observability
  • Nonlinear systems
  • Sliding mode observers
  • Uncertain systems

Cite this