Neural Network Modelling Applied for Model-Based Fault Detection

Zhanqun Shi, Yibo Fan, Fengshou Gu, Abdul Hannan Ali, Andrew Ball

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

This paper aims to combine neural network modelling with model-based fault detection. An accurate and robust model is critical in model-based fault detection. However, the development of such a model is the most difficult task especially when a non-linear system is involved. The problem comes not only from the lack of concerned information about model parameters, but also from the inevitable linearization. In order to solve this problem, neural networks are introduced in this paper. Instead of using conventional neural network modelling, the neural network is only used to approximate the non-linear part of the system, leaving the linear part to be represented by a mathematical model. This new scheme of integration between neural network and mathematical model (NNMM) allows the compensation of the error from conventional modelling methods. Simultaneously, it keeps the residual signatures physically interpretable.

LanguageEnglish
Title of host publicationASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004
PublisherAmerican Society of Mechanical Engineers (ASME)
Pages149-155
Number of pages7
Volume3
ISBN (Electronic)0791837416
ISBN (Print)0791841758
DOIs
Publication statusPublished - 19 Jul 2004
Externally publishedYes
Event7th Biennial Conference on Engineering Systems Design and Analysis - Manchester, United Kingdom
Duration: 19 Jul 200422 Jul 2004
Conference number: 7

Conference

Conference7th Biennial Conference on Engineering Systems Design and Analysis
CountryUnited Kingdom
CityManchester
Period19/07/0422/07/04

Fingerprint

Fault detection
Neural networks
Mathematical models
Linearization
Nonlinear systems

Cite this

Shi, Z., Fan, Y., Gu, F., Ali, A. H., & Ball, A. (2004). Neural Network Modelling Applied for Model-Based Fault Detection. In ASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004 (Vol. 3, pp. 149-155). [ESDA2004-58197] American Society of Mechanical Engineers (ASME). https://doi.org/10.1115/ESDA2004-58197
Shi, Zhanqun ; Fan, Yibo ; Gu, Fengshou ; Ali, Abdul Hannan ; Ball, Andrew. / Neural Network Modelling Applied for Model-Based Fault Detection. ASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004. Vol. 3 American Society of Mechanical Engineers (ASME), 2004. pp. 149-155
@inproceedings{6deaf8bd47244635b4a1cbdda3edefd0,
title = "Neural Network Modelling Applied for Model-Based Fault Detection",
abstract = "This paper aims to combine neural network modelling with model-based fault detection. An accurate and robust model is critical in model-based fault detection. However, the development of such a model is the most difficult task especially when a non-linear system is involved. The problem comes not only from the lack of concerned information about model parameters, but also from the inevitable linearization. In order to solve this problem, neural networks are introduced in this paper. Instead of using conventional neural network modelling, the neural network is only used to approximate the non-linear part of the system, leaving the linear part to be represented by a mathematical model. This new scheme of integration between neural network and mathematical model (NNMM) allows the compensation of the error from conventional modelling methods. Simultaneously, it keeps the residual signatures physically interpretable.",
keywords = "Control system, Fault detection, Model-based approach, Modelling, Neural networks",
author = "Zhanqun Shi and Yibo Fan and Fengshou Gu and Ali, {Abdul Hannan} and Andrew Ball",
year = "2004",
month = "7",
day = "19",
doi = "10.1115/ESDA2004-58197",
language = "English",
isbn = "0791841758",
volume = "3",
pages = "149--155",
booktitle = "ASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004",
publisher = "American Society of Mechanical Engineers (ASME)",
address = "United States",

}

Shi, Z, Fan, Y, Gu, F, Ali, AH & Ball, A 2004, Neural Network Modelling Applied for Model-Based Fault Detection. in ASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004. vol. 3, ESDA2004-58197, American Society of Mechanical Engineers (ASME), pp. 149-155, 7th Biennial Conference on Engineering Systems Design and Analysis, Manchester, United Kingdom, 19/07/04. https://doi.org/10.1115/ESDA2004-58197

Neural Network Modelling Applied for Model-Based Fault Detection. / Shi, Zhanqun; Fan, Yibo; Gu, Fengshou; Ali, Abdul Hannan; Ball, Andrew.

ASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004. Vol. 3 American Society of Mechanical Engineers (ASME), 2004. p. 149-155 ESDA2004-58197.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Neural Network Modelling Applied for Model-Based Fault Detection

AU - Shi, Zhanqun

AU - Fan, Yibo

AU - Gu, Fengshou

AU - Ali, Abdul Hannan

AU - Ball, Andrew

PY - 2004/7/19

Y1 - 2004/7/19

N2 - This paper aims to combine neural network modelling with model-based fault detection. An accurate and robust model is critical in model-based fault detection. However, the development of such a model is the most difficult task especially when a non-linear system is involved. The problem comes not only from the lack of concerned information about model parameters, but also from the inevitable linearization. In order to solve this problem, neural networks are introduced in this paper. Instead of using conventional neural network modelling, the neural network is only used to approximate the non-linear part of the system, leaving the linear part to be represented by a mathematical model. This new scheme of integration between neural network and mathematical model (NNMM) allows the compensation of the error from conventional modelling methods. Simultaneously, it keeps the residual signatures physically interpretable.

AB - This paper aims to combine neural network modelling with model-based fault detection. An accurate and robust model is critical in model-based fault detection. However, the development of such a model is the most difficult task especially when a non-linear system is involved. The problem comes not only from the lack of concerned information about model parameters, but also from the inevitable linearization. In order to solve this problem, neural networks are introduced in this paper. Instead of using conventional neural network modelling, the neural network is only used to approximate the non-linear part of the system, leaving the linear part to be represented by a mathematical model. This new scheme of integration between neural network and mathematical model (NNMM) allows the compensation of the error from conventional modelling methods. Simultaneously, it keeps the residual signatures physically interpretable.

KW - Control system

KW - Fault detection

KW - Model-based approach

KW - Modelling

KW - Neural networks

UR - http://www.scopus.com/inward/record.url?scp=8844225002&partnerID=8YFLogxK

U2 - 10.1115/ESDA2004-58197

DO - 10.1115/ESDA2004-58197

M3 - Conference contribution

SN - 0791841758

VL - 3

SP - 149

EP - 155

BT - ASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004

PB - American Society of Mechanical Engineers (ASME)

ER -

Shi Z, Fan Y, Gu F, Ali AH, Ball A. Neural Network Modelling Applied for Model-Based Fault Detection. In ASME 7th Biennial Conference on Engineering Systems Design and Analysis, ESDA 2004. Vol. 3. American Society of Mechanical Engineers (ASME). 2004. p. 149-155. ESDA2004-58197 https://doi.org/10.1115/ESDA2004-58197