Descriptor state‐bounding observer design for positive Markov jump linear systems with sensor faults: Simultaneous state and faults estimation

Summary This article presents a descriptor observer design approach for positive Markov jump linear systems subject to interval parameter uncertainties and sensor faults. First, by taking the sensor fault term as an auxiliary state, an augmented descriptor system is constructed. A pair of positive o...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:International journal of robust and nonlinear control Ročník 30; číslo 5; s. 2113 - 2129
Hlavní autoři: Song, Xiaoqi, Lam, James, Chen, Xiaoming, Zhu, Bohao
Médium: Journal Article
Jazyk:angličtina
Vydáno: Bognor Regis Wiley Subscription Services, Inc 25.03.2020
Témata:
Algorithms
descriptor system
Dynamic stability
fault estimation
Faults
Iterative methods
Linear systems
Markov jump linear system
Optimization
Parameter uncertainty
positive system
Robustness (mathematics)
Sensors
state‐bounding observer
ISSN:1049-8923, 1099-1239
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí:Summary This article presents a descriptor observer design approach for positive Markov jump linear systems subject to interval parameter uncertainties and sensor faults. First, by taking the sensor fault term as an auxiliary state, an augmented descriptor system is constructed. A pair of positive observers with state‐bounding feature is then proposed, which enables simultaneous estimation of the system state and sensor faults. A necessary and sufficient condition on existence of the desired state‐bounding observer is derived by considering positivity and robust mean exponential stability of corresponding observer error dynamics. An iterative optimization algorithm is developed for the computation of the optimized observer matrices. Finally, a numerical example is presented to show the validity of the proposed methods.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:1049-8923
1099-1239
DOI:10.1002/rnc.4870