Title :
Fault isolation using Model-on-Demand algorithm
Author :
Efendic, Hajrudin ; del Re, Luigi
Author_Institution :
Inst. for Design & Control of Mechatronical Syst., Johannes Kepler Univ., Linz, Austria
fDate :
Oct. 30 2013-Nov. 1 2013
Abstract :
This paper discusses the Model-on-Demand algorithm for fault isolation in large and complex technical systems. The main objective of the proposed algorithm is to improve fault isolation performance metrics by applying a structural analysis approach which targets the isolability problem in structurally time-invariant systems. This objective is achieved by the utilization of a reverse modeling approach in order to modify the structural matrix of the system in such a way to improve its isolability property. The proposed algorithm was tested in a real technical system and an overview of the experimental results is given here.
Keywords :
fault diagnosis; fault tolerant control; large-scale systems; complex technical systems; fault isolation performance metrics; isolability problem; isolability property; large technical systems; model-on-demand algorithm; reverse modeling approach; structural analysis approach; structural matrix; structurally time-invariant systems; Accuracy; Algorithm design and analysis; Automotive engineering; Engines; Fault detection; Fault diagnosis; Measurement; Model-on-Demand; complex technical systems; falt diagnosis; fault isolation; structural analysis;
Conference_Titel :
Information, Communication and Automation Technologies (ICAT), 2013 XXIV International Symposium on
Conference_Location :
Sarajevo
DOI :
10.1109/ICAT.2013.6684092