Remaining useful life prediction of aeroengine based on Parallel-GraphSAGE-GRU
Journal of Mechanical StrengthVol. 48, Issue 7, Pages: 32-39(2026)
作者机构:
1.郑州工业应用技术学院 机电工程学院,郑州 451150
2.河南省隧道工程机械工程技术研究中心,郑州 451150
3.河南理工大学 机械与动力工程学院,焦作 454003
作者简介:
GUO Wenwen, E-mail: gzh47702@163.com
基金信息:
National Natural Science Foundation of China(52404163);Science and Technology Plan Project of Henan Province(252102220079;252102221003);New Phase Henan Provincial Key Discipline (Mechanical Engineering) (EDU〔2023〕414)(教研〔2023〕414号);Henan Provincial Key Research Project in Higher Education Institutions(25B460020)
GUO Wenwen,WANG Guohu,YUAN Yongliang,et al. Remaining useful life prediction of aeroengine based on Parallel-GraphSAGE-GRU[J]. Journal of Mechanical Strength,2026,48(7):32-39.
GUO Wenwen,WANG Guohu,YUAN Yongliang,et al. Remaining useful life prediction of aeroengine based on Parallel-GraphSAGE-GRU[J]. Journal of Mechanical Strength,2026,48(7):32-39.DOI: 10.16579/j.issn.1001.9669.2026.07.004.
Remaining useful life prediction of aeroengine based on Parallel-GraphSAGE-GRU
Deep learning-based remaining useful life (RUL) prediction of aeroengine provides a foundation for guaranteeing the reliability of the aeroengine. However
most researches are devoted to the acquisition of degenerate information in Euclidean space
while topological relations in non-Euclidean space can provide additional degenerate information. Thus
a novel RUL prediction method named parallel graph node embedding learning-gated recurrent unit(Parallel-GraphSAGE-GRU) was proposed in this work.
Methods
2
Firstly
the maximum mutual information method was used to describe the relationships among monitoring parameters of aeroengine. Then
a series of graph inputs was constructed using a fixed time window to obtain topological relations in the non-Euclidean space of the aeroengine. Besides
GRU was used for the acquisition the temporal degradation relations of the monitoring parameters for aeroengine in the Euclidean space. The implicit vectors of GraphSAGE and GRU at each moment were concatenated and mapped to the RUL values of the input degradation information by fully connected neural network. The C-MAPSS dataset was used to validate the effectiveness and advancement of the proposed method
and the best
R
2
of the proposed method reached 0.
988 6.
Results
2
The results indicate that this method is feasible in ensuring prediction accuracy and computational efficiency.
Mu Hanxiao , Zheng Jianfei , Hu Changhua , et al . Remaining useful life prediction of multivariate degradation equipment based on CDBN and BiLSTM [J]. Acta Aeronautica et Astronautica Sinica , 2022 , 43 ( 7 ): 301 - 312 . (In Chinese)
Liu Yefeng , Wang Shuai , Liu Jingjing , et al . Bearing remaining life prediction based on TCN-SA and Bi-GRU [J]. Computer Integrated Manufacturing Systems , 2026 , 32 ( 4 ): 1428 - 1437 . (In Chinese)
Xi Xiaopeng , Chen Maoyin , Zhou Donghua . Remaining useful life prediction based on multirate sensor measurement data [J]. Control Engineering of China , 2020 , 27 ( 5 ): 771 - 775 . (In Chinese)
Zhang Luyihang , Yang Yanming , Chen Yongzhan , et al . Remaining useful life prediction of variable-operating turbofan engine based on VMD-CNN-BiLSTM [J]. Journal of Beijing University of Aeronautics and Astronautics , 2026 , 52 ( 4 ): 1279 - 1289 . (In Chinese)
Wu Jiajun , Su Chun , Zhang Yuru . Remaining useful life prediction based on double self-attention mechanism and long short-term memory network [J]. Systems Engineering and Electronics , 2024 , 46 ( 6 ): 1986 - 1994 . (In Chinese)
Lan Jie , Li Ning , Li Zhining , et al . Method of RUL prediction based on GRU Encoder-decoder and attentional mechanism [J]. Modern Electronics Technique , 2023 , 46 ( 8 ): 99 - 105 . (In Chinese)
Qiu Z W , Li W L , Tang T , et al . Denoising graph neural network based hydraulic component fault diagnosis method [J]. Mechanical Systems and Signal Processing , 2023 , 204 : 110828 .
Ye Xin , Wang Dongfeng , Qi Jianjun , et al . Research on residual life prediction of external gear pump based on IVMD and GNN [J]. Hydraulics Pneumatics & Seals , 2022 , 42 ( 8 ): 9 - 17 . (In Chinese)
Liu J L , Ong G P , Chen X Q . GraphSAGE-based traffic speed forecasting for segment network with sparse data [J]. IEEE Transactions on Intelligent Transportation Systems , 2022 , 23 ( 3 ): 1755 - 1766 .
Chai Tong , Yuan Yiping , Ma Junyan , et al . State prediction of wind turbine generator based on K-CNN and N-GRU [J]. Journal of Mechanical Strength , 2023 , 45 ( 5 ): 1043 - 1049 . (In Chinese)
Liu Qingjiang , Sun Chunliang , Zhou Jichang , et al . Fault warning technology for compressor unit based on integration of Attention and GRU [J]. Petroleum Engineering Construction , 2024 , 50 ( 4 ): 35 - 39 . (In Chinese)
Qin Zixuan , Zhang Xiaodong , Bai Guangzhi , et al . Aeroengine remaining useful life prediction based on multi-scale feature fusion [J]. Aeroengine , 2024 , 50 ( 4 ): 114 - 120 . (In Chinese)
Zhou J H , Qin Y , Luo J , et al . Remaining useful life prediction by distribution contact ratio health indicator and consolidated memory GRU [J]. IEEE Transactions on Industrial Informatics , 2023 , 19 ( 7 ): 8472 - 8483 .