On applicability of the interacting multiple-model approach to state estimation for systems with sojourn-time-dependent Markov model switching

• Autor: Petrov, A.I. ; Zubov, A.G.
• Assuntos: Artificial intelligence ; Bayesian methods ; Eigenvalues and eigenfunctions ; History ; Merging ; Robust stability ; State estimation ; Sufficient conditions ; Symmetric matrices ; Uncertain systems
• É parte de: IEEE transactions on automatic control, 1996-01, Vol.41 (1), p.136-140
• Notas: ObjectType-Article-2
  SourceType-Scholarly Journals-1
  ObjectType-Feature-1
  content type line 23
  
• Descrição: In this paper the attempt at the interacting multiple-model (IMM) method extension to the state estimation problem with semi-Markov [sojourn-time-dependent Markov (STDM)] system model switching is analyzed. It is demonstrated that such an STDM-IMM approach does not properly take into account the specific character of the STDM switching in the system, and it becomes a reason for the reduction of estimation accuracy. For this problem it is shown that a hypotheses merging should be restricted in such a way that not only the current system model, but also the sojourn time in the model, should be given in hypotheses to be tested. Some other aspects of Bayesian estimation in a switching environment are also discussed.
  
• Editor: New York, NY: IEEE
  
• Idioma: Inglês;Russo