Fault Identification for a Closed-Loop Control System Based on an Improved Deep Neural Network

Bowen Sun, Jiongqi Wang, Zhangming He, Haiyin Zhou, Fengshou Gu

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Fault identification for closed-loop control systems is a future trend in the field of fault diagnosis. Due to the inherent feedback adjustment mechanism, a closed-loop control system is generally very robust to external disturbances and internal noises. Closed-loop control systems often encourage faults to propagate inside the systems, which may lead to the consequence that faults amplitude becomes smaller and fault characteristics difference becomes more inapparent. Hence, it has been challenging to achieve fault identification for such systems. Traditional fault identification methods are not particularly designed for closed-loop control systems and thus cannot be applied directly. In this work, a new fault identification method is proposed, which is based on the deep neural network for closed-loop control systems. Firstly, the fault propagation mechanism in closed-loop control systems is theoretically derived, and the influence of fault propagation on system variables is analyzed. Then deep neural network is applied to find fault characteristics difference between different data modes, and a sliding window is used to amplify the fault-to-noise ratio and characteristics difference, with an aim to increase the identification performance. To verify this method, the simulations that are based on a numerical simulation model, the Tennessee industrial system and the satellite attitude control system are conducted. The results show that the proposed method is more feasible and more effective in fault identification for closed-loop control systems compared with traditional data-driven identification methods, including distance-based and angle-based identification methods
Original languageEnglish
Number of pages18
JournalSensors (Switzerland)
Volume19
Issue number9
DOIs
Publication statusPublished - 8 May 2019

Fingerprint

Closed loop control systems
Identification (control systems)
Noise
Social Adjustment
satellite attitude control
Attitude control
propagation
Deep neural networks
Failure analysis
sliding
Satellites
disturbances
simulation
Feedback
adjusting
Control systems
trends
Computer simulation

Cite this

Sun, Bowen ; Wang, Jiongqi ; He, Zhangming ; Zhou, Haiyin ; Gu, Fengshou. / Fault Identification for a Closed-Loop Control System Based on an Improved Deep Neural Network. In: Sensors (Switzerland). 2019 ; Vol. 19, No. 9.
@article{214e5b0f6a36417780486c8c6023382e,
title = "Fault Identification for a Closed-Loop Control System Based on an Improved Deep Neural Network",
abstract = "Fault identification for closed-loop control systems is a future trend in the field of fault diagnosis. Due to the inherent feedback adjustment mechanism, a closed-loop control system is generally very robust to external disturbances and internal noises. Closed-loop control systems often encourage faults to propagate inside the systems, which may lead to the consequence that faults amplitude becomes smaller and fault characteristics difference becomes more inapparent. Hence, it has been challenging to achieve fault identification for such systems. Traditional fault identification methods are not particularly designed for closed-loop control systems and thus cannot be applied directly. In this work, a new fault identification method is proposed, which is based on the deep neural network for closed-loop control systems. Firstly, the fault propagation mechanism in closed-loop control systems is theoretically derived, and the influence of fault propagation on system variables is analyzed. Then deep neural network is applied to find fault characteristics difference between different data modes, and a sliding window is used to amplify the fault-to-noise ratio and characteristics difference, with an aim to increase the identification performance. To verify this method, the simulations that are based on a numerical simulation model, the Tennessee industrial system and the satellite attitude control system are conducted. The results show that the proposed method is more feasible and more effective in fault identification for closed-loop control systems compared with traditional data-driven identification methods, including distance-based and angle-based identification methods",
keywords = "Closed-loop control system, Fault Diagnosis, Deep neural networks, Sliding window, identification",
author = "Bowen Sun and Jiongqi Wang and Zhangming He and Haiyin Zhou and Fengshou Gu",
year = "2019",
month = "5",
day = "8",
doi = "10.3390/s19092131",
language = "English",
volume = "19",
journal = "Sensors",
issn = "1424-3210",
publisher = "Multidisciplinary Digital Publishing Institute (MDPI)",
number = "9",

}

Fault Identification for a Closed-Loop Control System Based on an Improved Deep Neural Network. / Sun, Bowen; Wang, Jiongqi; He, Zhangming; Zhou, Haiyin; Gu, Fengshou.

In: Sensors (Switzerland), Vol. 19, No. 9, 08.05.2019.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Fault Identification for a Closed-Loop Control System Based on an Improved Deep Neural Network

AU - Sun, Bowen

AU - Wang, Jiongqi

AU - He, Zhangming

AU - Zhou, Haiyin

AU - Gu, Fengshou

PY - 2019/5/8

Y1 - 2019/5/8

N2 - Fault identification for closed-loop control systems is a future trend in the field of fault diagnosis. Due to the inherent feedback adjustment mechanism, a closed-loop control system is generally very robust to external disturbances and internal noises. Closed-loop control systems often encourage faults to propagate inside the systems, which may lead to the consequence that faults amplitude becomes smaller and fault characteristics difference becomes more inapparent. Hence, it has been challenging to achieve fault identification for such systems. Traditional fault identification methods are not particularly designed for closed-loop control systems and thus cannot be applied directly. In this work, a new fault identification method is proposed, which is based on the deep neural network for closed-loop control systems. Firstly, the fault propagation mechanism in closed-loop control systems is theoretically derived, and the influence of fault propagation on system variables is analyzed. Then deep neural network is applied to find fault characteristics difference between different data modes, and a sliding window is used to amplify the fault-to-noise ratio and characteristics difference, with an aim to increase the identification performance. To verify this method, the simulations that are based on a numerical simulation model, the Tennessee industrial system and the satellite attitude control system are conducted. The results show that the proposed method is more feasible and more effective in fault identification for closed-loop control systems compared with traditional data-driven identification methods, including distance-based and angle-based identification methods

AB - Fault identification for closed-loop control systems is a future trend in the field of fault diagnosis. Due to the inherent feedback adjustment mechanism, a closed-loop control system is generally very robust to external disturbances and internal noises. Closed-loop control systems often encourage faults to propagate inside the systems, which may lead to the consequence that faults amplitude becomes smaller and fault characteristics difference becomes more inapparent. Hence, it has been challenging to achieve fault identification for such systems. Traditional fault identification methods are not particularly designed for closed-loop control systems and thus cannot be applied directly. In this work, a new fault identification method is proposed, which is based on the deep neural network for closed-loop control systems. Firstly, the fault propagation mechanism in closed-loop control systems is theoretically derived, and the influence of fault propagation on system variables is analyzed. Then deep neural network is applied to find fault characteristics difference between different data modes, and a sliding window is used to amplify the fault-to-noise ratio and characteristics difference, with an aim to increase the identification performance. To verify this method, the simulations that are based on a numerical simulation model, the Tennessee industrial system and the satellite attitude control system are conducted. The results show that the proposed method is more feasible and more effective in fault identification for closed-loop control systems compared with traditional data-driven identification methods, including distance-based and angle-based identification methods

KW - Closed-loop control system

KW - Fault Diagnosis

KW - Deep neural networks

KW - Sliding window

KW - identification

U2 - 10.3390/s19092131

DO - 10.3390/s19092131

M3 - Article

VL - 19

JO - Sensors

JF - Sensors

SN - 1424-3210

IS - 9

ER -