Robust neural network fault estimation approach for nonlinear dynamic systems with applications to wind turbine systems

Reihane Rahimilarki; Zhiwei Gao; Aihua Zhang; Richard Binns

doi:10.1109/TII.2019.2893845

Robust neural network fault estimation approach for nonlinear dynamic systems with applications to wind turbine systems

Reihane Rahimilarki, Zhiwei Gao^*, Aihua Zhang, Richard Binns

^*Corresponding author for this work

Northumbria University

Research output: Contribution to journal › Article › peer-review

100 Citations (Scopus)

33 Downloads (Pure)

Abstract

In this paper, a robust fault estimation approach is proposed for multi-input and multi-output nonlinear dynamic systems on the basis of back propagation neural networks. The augmented system approach, input-to-state stability theory, linear matrix inequality optimization, and neural network training/learning are integrated so that a robust simultaneous estimate of system states and actuator faults are achieved. The proposed approaches are finally applied to a 4.8 MW wind turbine benchmark system, and the effectiveness is well demonstrated.

Original language	English
Article number	8616801
Pages (from-to)	6302-6312
Number of pages	11
Journal	IEEE Transactions on Industrial Informatics
Volume	15
Issue number	12
Early online date	17 Jan 2019
DOIs	https://doi.org/10.1109/TII.2019.2893845
Publication status	Published - 5 Dec 2019

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

Artificial neural network (ANN)
fault estimation
input-to-state stability
linear matrix inequality
robustness
wind turbine systems

Access to Document

10.1109/TII.2019.2893845

ALL_TII-18-2817Accepted author manuscript, 2.93 MB

Cite this

@article{b87324e9b5d2401eb380146276771ae0,

title = "Robust neural network fault estimation approach for nonlinear dynamic systems with applications to wind turbine systems",

abstract = "In this paper, a robust fault estimation approach is proposed for multi-input and multi-output nonlinear dynamic systems on the basis of back propagation neural networks. The augmented system approach, input-to-state stability theory, linear matrix inequality optimization, and neural network training/learning are integrated so that a robust simultaneous estimate of system states and actuator faults are achieved. The proposed approaches are finally applied to a 4.8 MW wind turbine benchmark system, and the effectiveness is well demonstrated.",

keywords = "Artificial neural network (ANN), fault estimation, input-to-state stability, linear matrix inequality, robustness, wind turbine systems",

author = "Reihane Rahimilarki and Zhiwei Gao and Aihua Zhang and Richard Binns",

note = "Research funded by National Natural Science Foundation of China (61673074), American Friends of the Alexander von Humboldt Foundation (GRO/1117303), E\&E Faculty, University of Northumbria",

year = "2019",

month = dec,

day = "5",

doi = "10.1109/TII.2019.2893845",

language = "English",

volume = "15",

pages = "6302--6312",

journal = "IEEE Transactions on Industrial Informatics",

issn = "1551-3203",

publisher = "IEEE",

number = "12",

}

TY - JOUR

T1 - Robust neural network fault estimation approach for nonlinear dynamic systems with applications to wind turbine systems

AU - Rahimilarki, Reihane

AU - Gao, Zhiwei

AU - Zhang, Aihua

AU - Binns, Richard

N1 - Research funded by National Natural Science Foundation of China (61673074), American Friends of the Alexander von Humboldt Foundation (GRO/1117303), E&E Faculty, University of Northumbria

PY - 2019/12/5

Y1 - 2019/12/5

N2 - In this paper, a robust fault estimation approach is proposed for multi-input and multi-output nonlinear dynamic systems on the basis of back propagation neural networks. The augmented system approach, input-to-state stability theory, linear matrix inequality optimization, and neural network training/learning are integrated so that a robust simultaneous estimate of system states and actuator faults are achieved. The proposed approaches are finally applied to a 4.8 MW wind turbine benchmark system, and the effectiveness is well demonstrated.

AB - In this paper, a robust fault estimation approach is proposed for multi-input and multi-output nonlinear dynamic systems on the basis of back propagation neural networks. The augmented system approach, input-to-state stability theory, linear matrix inequality optimization, and neural network training/learning are integrated so that a robust simultaneous estimate of system states and actuator faults are achieved. The proposed approaches are finally applied to a 4.8 MW wind turbine benchmark system, and the effectiveness is well demonstrated.

KW - Artificial neural network (ANN)

KW - fault estimation

KW - input-to-state stability

KW - linear matrix inequality

KW - robustness

KW - wind turbine systems

UR - https://www.scopus.com/pages/publications/85067456293

U2 - 10.1109/TII.2019.2893845

DO - 10.1109/TII.2019.2893845

M3 - Article

AN - SCOPUS:85067456293

SN - 1551-3203

VL - 15

SP - 6302

EP - 6312

JO - IEEE Transactions on Industrial Informatics

JF - IEEE Transactions on Industrial Informatics

IS - 12

M1 - 8616801

ER -

Robust neural network fault estimation approach for nonlinear dynamic systems with applications to wind turbine systems

Abstract

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this