An application of decentralized estimation in a fault detection problem

This paper presents a design of a decentralized fault detection and isolation (FDI) filter by means of an overlapping decentralized estimation algorithm based on a consensus strategy. An efficient solution to the FDI problem can be obtained by an adequate system decomposition into overlapping subsys...

Full description

Bibliographic Details
Main Authors: Tadić Predrag R., Stanković Miloš S., Stanković Srđan S., Đurović Željko M.
Format: Article
Language:English
Published: Faculty of Technical Sciences in Cacak 2009-01-01
Series:Serbian Journal of Electrical Engineering
Subjects:
Online Access:http://www.doiserbia.nb.rs/img/doi/1451-4869/2009/1451-48690903373T.pdf
Description
Summary:This paper presents a design of a decentralized fault detection and isolation (FDI) filter by means of an overlapping decentralized estimation algorithm based on a consensus strategy. An efficient solution to the FDI problem can be obtained by an adequate system decomposition into overlapping subsystems and the construction of local FDI estimators aimed at achieving the desired performance. The general aspects and properties of a consensus based estimator are described in the first part of the paper. An applicability of such an estimator to an FDI problem in a large scale system is discussed next. Namely, a case study related to the detection of fire dissymmetry in a thermal power plant boiler is presented, including the process description and identification procedure, comparison between the results obtained by local and decentralized estimators and conclusions concerning their validity.
ISSN:1451-4869
2217-7183