A fault diagnosis method that combines multi-source signal fusion with generative adversarial network (GAN) is proposed to establish a fault diagnosis model. Principal component analysis (PCA) is initially utilized to fuse vibration signals from multiple sensors to obtain more reliable, precise, and comprehensive fault signal features. Subsequently, the Markov transition field (MTF) is used to convert one-dimensional vibration signals into two-dimensional image outputs, preserving the time-dependency of the signals. Finally, these two-dimensional image data serve as input for the GAN model, which consists of a generator and a discriminator. The adversarial interaction between these components facilitates the recognition of fault patterns. The performance of the fault diagnosis model is validated using a public bearing fault dataset, achieving a 95.5% accuracy rate with minimal training samples, demonstrating the feasibility of the model.
ZhaoQQZhangYYYangY. Multi-layers fault diagnosis method based on multi-source information. In: Proceedings of 2013 IEEE international conference on vehicular electronics and safety, Dongguan, China, 28–30 July 2013, pp.189–193. New York: IEEE.
2.
SongMBZhiYFAnMD, et al. Centrifugal pump cavitation fault diagnosis based on feature-level multi-source information fusion. Processes2024; 12(1): 196.
3.
ZhaoYY.A review of fault diagnosis methods based on machine learning. China Equip Eng2024; 1: 173–175.
4.
YinWHWangFSunQ, et al. A review on the application of artificial intelligence in fault diagnosis of marine diesel engines. Ship Sci Technol2023; 45(24): 122–127.
5.
XuZKGaoXYJunF, et al. A novel fault diagnosis method under limited samples based on an extreme learning machine and meta-learning. J Taiwan Inst Chem Eng2024; 161: 105522.
KumarPTSaimuruganMHaranRH, et al. A multi-sensor information fusion for fault diagnosis of a gearbox utilizing discrete wavelet features. Meas Sci Technol2019; 30(8): 085101.
8.
WangDCLiYBJiaL, et al. Novel three-stage feature fusion method of multimodal data for bearing fault diagnosis. IEEE Trans Instrum Meas2021; 70: 3514710.
9.
BuchaiahSShakyaP.Bearing fault diagnosis and prognosis using data fusion based feature extraction and feature selection. Measurement2022; 188: 110506.
10.
HanDYTianJHXueP, et al. A novel intelligent fault diagnosis method based on a dual convolutional neural network with multi-level information fusion. J Mech Sci Technol2021; 35(8): 3331–3345.
11.
TangYXiaoXGuanMT, et al. Fault diagnosis method for reciprocating compressors based on multi-source signal fusion. J Shandong Univ (Eng Sci)2023; 53(5): 10–19.
12.
YangGXChenXHYuHK.Fault diagnosis method for rolling bearings based on multi-source signal fusion and SE-CNN. Coal Mine Machinery2024; 45(4): 158–160.
13.
YangGChenCHGuoCY.Genetic algorithm for solving the vanishing gradient problem in deep neural network. Basic Clin Pharmacol Toxicol2021; 128: 82–83.
14.
JungyeopOSungkyuKChanghyeonL, et al. Preventing vanishing gradient problem of hardware neuromorphic system by implementing imidazole-based memristive ReLU activation neuron. Adv Mater2023; 35(24): e2300023.
15.
XueTLGengJ.Fault diagnosis of single-phase grounding in distribution networks based on optimized deep neural networks. Electr Autom2021; 43(1): 88–91.
16.
TongJWHuangTLLongTG, et al. Research on multi-sensor data fusion algorithm based on principal component analysis. Microcontroller Embed Syst Appl2013; 13(5): 17–20.
17.
HanWLMaoDJCaiY, et al. Research on fault early warning for wind turbine pitch systems based on data fusion. Acta Energiae Solaris Sinica2022; 43(12): 236–241.
18.
XuZChenXLiY, et al. Hybrid multimodal feature fusion with multi-sensor for bearing fault diagnosis. Sensors2024; 24(6): 1792.
19.
ManasPSupriyoSKasturiG, et al. Multi-source data fusion technique for parametric fault diagnosis in analog circuits. Integration2022; 84: 92–101.
20.
WangMJWangWJZhangXN, et al. A new fault diagnosis of rolling bearing based on Markov transition field and CNN. Entropy2022; 24(6): 751.
21.
LeiCLXueLLJiaoMX, et al. Rolling bearing fault diagnosis by Markov transition field and multi-dimension convolutional neural network. Meas Sci Technol2022; 33(11): 114009.
22.
ChoiHSParkKJAnD, et al. Fault diagnosis method for human coexistence robots based on convolutional neural networks using time-series data generation and image encoding. Sensors2023; 23(24): 9753.
23.
GoodfellowIJPouget-AbadieJMirzaM, et al. Generative adversarial nets. In: Conference on neural information processing systems, Montreal, QC, Canada, 2014, pp.2672–2680. Red Hook, NY: Curran Associates, Inc.
