Fault estimation using algebraic observers

Guerra, Rafael Martínez; Luviano‐Juárez, Alberto; Rincon-Pasaye, J.J.

doi:10.1109/acc.2007.4282951

Cited by 5 publications

(1 citation statement)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Engineering sciences rely heavily on estimates of system states and observer diagnosis is used for estimating actuator and sensor faults [1][2][3][4]. Usually industrial systems are governed by a number of physical phenomena and various technology components; for this, the bond graph approach based on energy analysis and multiphysics is well suited.…”

Section: Introductionmentioning

confidence: 99%

Diagnosis of reverse osmosis desalination water system using bond graph approach

Sallami¹,

Mzoughi²,

Mami³

2018

Turk J Elec Eng & Comp Sci

View full text Add to dashboard Cite

Industrial systems have become quite complicated. It is therefore necessary to look for a reliable supervision system to properly treat the information and make the appropriate decision to stop the system or leave it in operation.Supervision of automated systems goes to the heart of the matter. This paper concerns the design of a new methodology for diagnosis based on bond graph modeling. We apply a diagnostic method by Luenberger observer using the bond graph approach on a desalination unit. Tests are carried out on the water input pressure delivered by the pump and the output flow rates of the reverse osmosis. This method will allow us to test normal and failure operations of an industrial system. The roles of diagnostic systems for industrial processes consist of detecting and locating faults, also called residues, that will affect these processes. In order to be effective, the diagnostic system should be robust for residue analysis and insensitive to false alarms. In this context, the development of the proposed method is described from the step of modeling to the step of generation of certain residues.

show abstract

Section: Introductionmentioning

confidence: 99%