A gray-box neural network-based model identification and fault estimation scheme for nonlinear dynamic systems

Zhaohui Cen, Jiaolong Wei, Rui Jiang

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

A novel gray-box neural network model (GBNNM), including multi-layer perception (MLP) neural network (NN) and integrators, is proposed for a model identification and fault estimation (MIFE) scheme. With the GBNNM, both the nonlinearity and dynamics of a class of nonlinear dynamic systems can be approximated. Unlike previous NN-based model identification methods, the GBNNM directly inherits system dynamics and separately models system nonlinearities. This model corresponds well with the object system and is easy to build. The GBNNM is embedded online as a normal model reference to obtain the quantitative residual between the object system output and the GBNNM output. This residual can accurately indicate the fault offset value, so it is suitable for differing fault severities. To further estimate the fault parameters (FPs), an improved extended state observer (ESO) using the same NNs (IESONN) from the GBNNM is proposed to avoid requiring the knowledge of ESO nonlinearity. Then, the proposed MIFE scheme is applied for reaction wheels (RW) in a satellite attitude control system (SACS). The scheme using the GBNNM is compared with other NNs in the same fault scenario, and several partial loss of effect (LOE) faults with different severities are considered to validate the effectiveness of the FP estimation and its superiority.

Original languageEnglish
Article number1350025
JournalInternational Journal of Neural Systems
Volume23
Issue number6
DOIs
Publication statusPublished - 1 Dec 2013
Externally publishedYes

Fingerprint

Identification (control systems)
Dynamical systems
Neural networks
Control nonlinearities
Attitude control
Parameter estimation
Wheels
Satellites
Control systems

Keywords

  • extended state observer
  • gray-box neural-network model
  • Model identification and fault estimation
  • nonlinear dynamic systems
  • reaction wheel

ASJC Scopus subject areas

  • Computer Networks and Communications

Cite this

A gray-box neural network-based model identification and fault estimation scheme for nonlinear dynamic systems. / Cen, Zhaohui; Wei, Jiaolong; Jiang, Rui.

In: International Journal of Neural Systems, Vol. 23, No. 6, 1350025, 01.12.2013.

Research output: Contribution to journalArticle

@article{b4a79c913f574ed5a522d5dd641deed4,
title = "A gray-box neural network-based model identification and fault estimation scheme for nonlinear dynamic systems",
abstract = "A novel gray-box neural network model (GBNNM), including multi-layer perception (MLP) neural network (NN) and integrators, is proposed for a model identification and fault estimation (MIFE) scheme. With the GBNNM, both the nonlinearity and dynamics of a class of nonlinear dynamic systems can be approximated. Unlike previous NN-based model identification methods, the GBNNM directly inherits system dynamics and separately models system nonlinearities. This model corresponds well with the object system and is easy to build. The GBNNM is embedded online as a normal model reference to obtain the quantitative residual between the object system output and the GBNNM output. This residual can accurately indicate the fault offset value, so it is suitable for differing fault severities. To further estimate the fault parameters (FPs), an improved extended state observer (ESO) using the same NNs (IESONN) from the GBNNM is proposed to avoid requiring the knowledge of ESO nonlinearity. Then, the proposed MIFE scheme is applied for reaction wheels (RW) in a satellite attitude control system (SACS). The scheme using the GBNNM is compared with other NNs in the same fault scenario, and several partial loss of effect (LOE) faults with different severities are considered to validate the effectiveness of the FP estimation and its superiority.",
keywords = "extended state observer, gray-box neural-network model, Model identification and fault estimation, nonlinear dynamic systems, reaction wheel",
author = "Zhaohui Cen and Jiaolong Wei and Rui Jiang",
year = "2013",
month = "12",
day = "1",
doi = "10.1142/S0129065713500251",
language = "English",
volume = "23",
journal = "International Journal of Neural Systems",
issn = "0129-0657",
publisher = "World Scientific Publishing Co. Pte Ltd",
number = "6",

}

TY - JOUR

T1 - A gray-box neural network-based model identification and fault estimation scheme for nonlinear dynamic systems

AU - Cen, Zhaohui

AU - Wei, Jiaolong

AU - Jiang, Rui

PY - 2013/12/1

Y1 - 2013/12/1

N2 - A novel gray-box neural network model (GBNNM), including multi-layer perception (MLP) neural network (NN) and integrators, is proposed for a model identification and fault estimation (MIFE) scheme. With the GBNNM, both the nonlinearity and dynamics of a class of nonlinear dynamic systems can be approximated. Unlike previous NN-based model identification methods, the GBNNM directly inherits system dynamics and separately models system nonlinearities. This model corresponds well with the object system and is easy to build. The GBNNM is embedded online as a normal model reference to obtain the quantitative residual between the object system output and the GBNNM output. This residual can accurately indicate the fault offset value, so it is suitable for differing fault severities. To further estimate the fault parameters (FPs), an improved extended state observer (ESO) using the same NNs (IESONN) from the GBNNM is proposed to avoid requiring the knowledge of ESO nonlinearity. Then, the proposed MIFE scheme is applied for reaction wheels (RW) in a satellite attitude control system (SACS). The scheme using the GBNNM is compared with other NNs in the same fault scenario, and several partial loss of effect (LOE) faults with different severities are considered to validate the effectiveness of the FP estimation and its superiority.

AB - A novel gray-box neural network model (GBNNM), including multi-layer perception (MLP) neural network (NN) and integrators, is proposed for a model identification and fault estimation (MIFE) scheme. With the GBNNM, both the nonlinearity and dynamics of a class of nonlinear dynamic systems can be approximated. Unlike previous NN-based model identification methods, the GBNNM directly inherits system dynamics and separately models system nonlinearities. This model corresponds well with the object system and is easy to build. The GBNNM is embedded online as a normal model reference to obtain the quantitative residual between the object system output and the GBNNM output. This residual can accurately indicate the fault offset value, so it is suitable for differing fault severities. To further estimate the fault parameters (FPs), an improved extended state observer (ESO) using the same NNs (IESONN) from the GBNNM is proposed to avoid requiring the knowledge of ESO nonlinearity. Then, the proposed MIFE scheme is applied for reaction wheels (RW) in a satellite attitude control system (SACS). The scheme using the GBNNM is compared with other NNs in the same fault scenario, and several partial loss of effect (LOE) faults with different severities are considered to validate the effectiveness of the FP estimation and its superiority.

KW - extended state observer

KW - gray-box neural-network model

KW - Model identification and fault estimation

KW - nonlinear dynamic systems

KW - reaction wheel

UR - http://www.scopus.com/inward/record.url?scp=84887033105&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84887033105&partnerID=8YFLogxK

U2 - 10.1142/S0129065713500251

DO - 10.1142/S0129065713500251

M3 - Article

VL - 23

JO - International Journal of Neural Systems

JF - International Journal of Neural Systems

SN - 0129-0657

IS - 6

M1 - 1350025

ER -