Open AccessMemory scheduling robust H ∞ filter‐based fault detection for discrete‐time polytopic uncertain systems over fading channels
Author(s) -
Feng Jian,
Han Kezhen,
Zhao Qing
Publication year - 2017
Publication title -
iet control theory and applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.059
H-Index - 108
eISSN - 1751-8652
pISSN - 1751-8644
DOI - 10.1049/iet-cta.2016.1248
Subject(s) - control theory (sociology) , fading , robustness (evolution) , fault detection and isolation , residual , linear matrix inequality , computer science , discrete time and continuous time , scheduling (production processes) , algorithm , mathematics , mathematical optimization , actuator , biochemistry , chemistry , decoding methods , statistics , control (management) , artificial intelligence , gene
A novel memory scheduling robust H ∞ fault detection filter (FDF) is proposed for a class of discrete‐time polytopic uncertain systems with fading channel communication networks. The main merit of this filter‐based fault detection method is that it can significantly improve the robustness of FDF to attenuate influences from external disturbance, channel fading and model uncertainty on fault detection accuracy. Designing such FDF involves three main stages. First of all, a memory scheduling FDF structure is proposed based on the utilisation of weighted historical filter's states over interval instants, and a residual error system is formulated based on time partition and state augmented approaches. Then, the parameter‐dependent Lyapunov method is further utilised to analyse the stochastic stability of the residual error system with the help of Finsler equivalent transformation. In the following, a two‐stage optimisation algorithm combined with scalar parameters method is constructed to design memory scheduling FDF in a less conservative linear matrix inequality manner. Finally, a random numerical verification with 300 test systems and a case study of an industrial continuous‐stirred tank reactor are exploited to show the effectiveness of obtained results.