Fault Detection of Aero-Engine Sensor Based on Inception-CNN

被引:26
作者
Du, Xiao [1 ]
Chen, Jiajie [1 ]
Zhang, Haibo [1 ]
Wang, Jiqiang [2 ]
机构
[1] Nanjing Univ Aeronaut & Astronaut, Coll Energy & Power Engn, Nanjing 210016, Peoples R China
[2] Chinese Acad Sci, Ningbo Inst Mat Technol & Engn, Ningbo 315201, Peoples R China
关键词
CNN; sensor fault detection; aircraft engine; Monte Carlo simulation method; inception block; DIAGNOSIS; ARCHITECTURE; MODEL;
D O I
10.3390/aerospace9050236
中图分类号
V [航空、航天];
学科分类号
08 ; 0825 ;
摘要
The aero-engine system is complex, and the working environment is harsh. As the fundamental component of the aero-engine control system, the sensor must monitor its health status. Traditional sensor fault detection algorithms often have many parameters, complex architecture, and low detection accuracy. Aiming at this problem, a convolutional neural network (CNN) whose basic unit is an inception block composed of convolution kernels of different sizes in parallel is proposed. The network fully extracts redundant analytical information between sensors through different size convolution kernels and uses it for aero-engine sensor fault detection. On the sensor failure dataset generated by the Monte Carlo simulation method, the detection accuracy of Inception-CNN is 95.41%, which improves the prediction accuracy by 17.27% and 12.69% compared with the best-performing non-neural network algorithm and simple BP neural networks tested in the paper, respectively. In addition, the method simplifies the traditional fault detection unit composed of multiple fusion algorithms into one detection algorithm, which reduces the complexity of the algorithm. Finally, the effectiveness and feasibility of the method are verified in two aspects of the typical sensor fault detection effect and fault detection and isolation process.
引用
收藏
页数:18
相关论文
共 38 条
[1]  
[艾剑良 Ai Jianliang], 2018, [中国科学. 技术科学, Scientia Sinica Technologica], V48, P326
[2]  
Baocheng H., 2001, J PROPULS TECHNOL, V22, P364, DOI DOI 10.13675/J.CNKI.TJJS.2001.05.004
[3]  
Cai Kai-long, 2008, Journal of Aerospace Power, V23, P1118
[4]   Fusing physics-based and deep learning models for prognostics [J].
Chao, Manuel Arias ;
Kulkarni, Chetan ;
Goebel, Kai ;
Fink, Olga .
RELIABILITY ENGINEERING & SYSTEM SAFETY, 2022, 217
[5]  
Chapman J., 2014, TOOLBOX MODELING ANA
[6]  
Chen C, 2012, CHIN CONTR CONF, P2071
[7]  
[崔建国 Cui Jianguo], 2018, [火力与指挥控制, Fire Control & Command Control], V43, P113
[8]   From Model, Signal to Knowledge: A Data-Driven Perspective of Fault Detection and Diagnosis [J].
Dai, Xuewu ;
Gao, Zhiwei .
IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2013, 9 (04) :2226-2238
[9]   BiGRU-CNN Neural Network Applied to Electric Energy Theft Detection [J].
Duarte Soares, Lucas ;
Queiroz, Altamira de Souza ;
Lopez, Gloria P. ;
Carreno-Franco, Edgar M. ;
Lopez-Lezama, Jesus M. ;
Munoz-Galeano, Nicolas .
ELECTRONICS, 2022, 11 (05)
[10]   Single-Sensor Acoustic Emission Source Localization in Plate-Like Structures Using Deep Learning [J].
Ebrahimkhanlou, Arvin ;
Salamone, Salvatore .
AEROSPACE, 2018, 5 (02)