Fault diagnosis of servo valve based on multi-source signal and hybrid attention

doi:10.3785/j.issn.1008-973X.2026.09.002

Journal of ZheJiang University (Engineering Science)

2026, Vol. 60

Issue (9): 1851-1861 DOI: 10.3785/j.issn.1008-973X.2026.09.002

Fault diagnosis of servo valve based on multi-source signal and hybrid attention

Chen YANG1,3(

),Jianwen YAN1,3,*(

),Lei LI2,Guishan LI3

1. School of Mechatronics Engineering, Anhui University of Science and Technology, Huainan 232001, China
2. Institute of Green Design and Manufacturing Engineering, Hefei University of Technology, Hefei 230009, China
3. Hefei Metalforming Intelligent Manufacturing Limited Company, Hefei 230601, China

Download:

HTML

PDF(8012KB) HTML
Export: BibTeX | EndNote (RIS)

Abstract

A fault diagnosis method for electro-hydraulic servo valve based on multi-source signal fusion and hybrid attention mechanism was analyzed in order to improve the diagnostic accuracy and robustness for concealed leakage and blockage fault in electro-hydraulic servo valve under heavy-load, time-varying-speed and strong-noise condition, aiming to enhance the capability of the model for weak fault feature representation and recognition. A dual-channel multiscale Transformer fusion network for pressure and control signal, named DC-MTF-Net, was constructed oriented toward engineering application scenario. A convolutional attention module was employed to enhance multiscale temporal feature extraction, while cross-attention gated fusion was introduced to realize adaptive weighting of multi-source feature. A two-stage learning framework combining autoencoder pretraining and classification fine-tuning was adopted to improve feature representation and generalization capability in complex production environment. Multi-fault experiments conducted on an electro-hydraulic servo valve test platform under heavy-load and time-varying-speed condition showed that the proposed method improved diagnostic accuracy by at least 6%, compared with baseline models. Higher stability and robustness were demonstrated in comparative experiments involving single-channel diagnosis and various noise disturbance, verifying the effectiveness of the method in engineering application.

Key words： heavy-load and time-varying-speed condition electro-hydraulic servo valve multi-source signal fusion hybrid attention cross-attention gated fusion

Received: 28 October 2025 Published: 20 July 2026

CLC:

TP 181

Fund: 国家重点研发计划资助项目（2023YEB3210805，2024YFE03120000）；重点实验室开放课题资助项目（PA2024GDSK0066）.

Corresponding Authors: Jianwen YAN E-mail: 2023100049@aust.edu.cn;yanjw_hfm@163.com

	Service
	E-mail this article
	Add to my bookshelf
	Add to citation manager
	E-mail Alert
	RSS
	Articles by authors
	Chen YANG
	Jianwen YAN
	Lei LI
	Guishan LI

Cite this article:

Chen YANG,Jianwen YAN,Lei LI,Guishan LI. Fault diagnosis of servo valve based on multi-source signal and hybrid attention. Journal of ZheJiang University (Engineering Science), 2026, 60(9): 1851-1861.

URL:

https://www.zjujournals.com/eng/10.3785/j.issn.1008-973X.2026.09.002 OR https://www.zjujournals.com/eng/Y2026/V60/I9/1851

基于多源信号与混合注意力的伺服阀故障诊断

为了提高重载与时变速及强噪声工况下电液伺服阀泄漏和阻塞隐蔽故障的诊断精度与鲁棒性，开展基于多源信号融合与混合注意力机制的电液伺服阀故障诊断方法研究，以增强模型对弱故障特征的表征与识别能力. 面向工程应用场景，构建压力与控制信号的双通道多尺度 Transformer 融合网络(DC-MTF-Net)，通过卷积注意力模块增强多尺度时序特征提取，引入交叉注意力门控融合实现多源特征的自适应加权. 采用自编码预训练与分类微调的两阶段学习框架，提升复杂生产环境下的特征表征与泛化能力. 基于重载与时变速的电液伺服阀试验平台多故障实验表明，与对比模型相比，所提方法的诊断准确率至少提高6%，在单通道诊断及多种噪声干扰对比实验下均表现出更高的稳定性和鲁棒性，验证了该方法在工程应用中的有效性.

关键词： 重载与时变速工况, 电液伺服阀, 多源信号融合, 混合注意力, 交叉注意力门控融合

Fig.1 Fault type of electro-hydraulic servo valve in engineering practice

Fig.2 Operating principle of servo valve blockage and leakage fault

Tab.1 Failure type of electro-hydraulic servo valve

Fig.3 Structure of DC-MTF-Net model

Fig.4 Cross-attention gated fusion

Fig.5 Main configuration of servo valve fault test bench

Tab.2 Main component and key parameter of test bench

Fig.6 Acquisition and transmission of control and pressure signal

Fig.7 Local dataset of control and pressure signal

Tab.3 Parameter of each layer in DC-MTF-Net model

Tab.4 Comparison of evaluation indicator for ablation experiment

Fig.8 Model output visualization under average metrics

Tab.5 Comparison of four evaluation metrics between DC-MTF-Net model and advanced models

Fig.9 Confusion matrix for single-signal and feature-fusion strategy

Tab.6 Model diagnostic accuracy and standard deviation under different noise intensity


[1]	陈俊翔, 姜宏达, 孔祥东, 等考虑时滞因素的负载口独立系统模式切换平稳性研究[J]. 中国机械工程, 2025, 36 (3): 414- 425 CHEN Junxiang, JIANG Hongda, KONG Xiangdong, et al Research on smoothness of mode switch of independent metering systems considering time-delay factors[J]. China Mechanical Engineering, 2025, 36 (3): 414- 425 doi: 10.3969/j.issn.1004-132X.2025.03.005

[2]	REN T, YIN Y, WANG X, et al Leakage mechanism analysis of electro-hydraulic servo valve under transverse vibration environment[J]. Engineering Failure Analysis, 2025, 169: 109181 doi: 10.1016/j.engfailanal.2024.109181

[3]	BRANCO RAMOS FILHO J R, DE NEGRI V J. Model-based fault detection for hydraulic servo proportional valves [C]//13th Scandinavian International Conference on Fluid Power. Linköping: Linköping University Electronic Press, 2013: 389−398.

[4]	LEI Y, YANG B, JIANG X, et al Applications of machine learning to machine fault diagnosis: a review and roadmap[J]. Mechanical Systems and Signal Processing, 2020, 138: 106587 doi: 10.1016/j.ymssp.2019.106587

[5]	胡渊豪, 宋艺博, 刘家辉, 等基于注意力卷积胶囊网络的电液比例伺服阀故障诊断[J]. 航空学报, 2024, 45 (15): 122- 131 HU Yuanhao, SONG Yibo, LIU Jiahui, et al Fault diagnosis of electro-hydraulic proportional servo valves based on attention convolutional capsule networks[J]. Acta Aeronautica et Astronautica Sinica, 2024, 45 (15): 122- 131 doi: 10.7527/S1000-6893.2024.30407

[6]	孙炜, 刘恒, 陶建峰, 等基于IndRNN-1DLCNN的负载口独立控制阀控缸系统故障诊断[J]. 浙江大学学报: 工学版, 2023, 57 (10): 2028- 2041 SUN Wei, LIU Heng, TAO Jianfeng, et al IndRNN-1DLCNN based fault diagnosis of independent metering valve-controlled hydraulic cylinder system[J]. Journal of Zhejiang University: Engineering Science, 2023, 57 (10): 2028- 2041 doi: 10.3785/j.issn.1008-973X.2023.10.012

[7]	赵洪利, 杨佳强基于融合卷积Transformer的航空发动机故障诊断[J]. 北京航空航天大学学报, 2025, 51 (4): 1117- 1126 ZHAO Hongli, YANG Jiaqiang Aero-engine fault diagnosis based on fusion convolutional Transformer[J]. Journal of Beijing University of Aeronautics and Astronautics, 2025, 51 (4): 1117- 1126

[8]	CHEN R, CHEN C, KELIMU M Hydraulic valve fault diagnosis based on EEWT and POA-KELM[J]. Measurement Science and Technology, 2025, 36 (1): 016147 doi: 10.1088/1361-6501/ad86d4

[9]	LI K, SUN Z, JIN H, et al Semi-supervised diagnosis method of refrigeration compressor hidden defect based on convolutional transformer autoencoder model[J]. International Journal of Refrigeration, 2024, 158: 47- 57 doi: 10.1016/j.ijrefrig.2023.10.021

[10]	LIU Z, NING H, WU M, et al A comprehensive diagnosis method of valve leakage faults based on bi-sensor information fusion[J]. Structural Health Monitoring, 2024, 23 (1): 512- 526 doi: 10.1177/14759217231174369

[11]	SHI J C, REN Y, TANG H S, et al Hydraulic directional valve fault diagnosis using a weighted adaptive fusion of multi-dimensional features of a multi-sensor[J]. Journal of Zhejiang University: Science A, 2022, 23 (4): 257- 271 doi: 10.1631/jzus.A2100394

[12]	陈俊英, 席月芸, 李朝阳多尺度局部特征和Transformer全局学习融合的发动机剩余寿命预测[J]. 自动化学报, 2024, 50 (9): 1818- 1830 CHEN Junying, XI Yueyun, LI Zhaoyang Prediction of aeroengine remaining life by combining multi-scale local features and transformer global learning[J]. Acta Automatica Sinica, 2024, 50 (9): 1818- 1830 doi: 10.16383/j.aas.c230634

[13]	WOO S, PARK J, LEE J Y, et al. CBAM: convolutional block attention module [M/OL]. [S. l.]: Springer, 2018: 11211 [2025-06-21]. https://link.springer.com/10.1007/978-3-030-01234-2_1.

[14]	谢雯, 王若男, 羊鑫, 等融合深度可分离卷积的多尺度残差UNet在PolSAR地物分类中的研究[J]. 电子与信息学报, 2023, 45 (8): 2975- 2985 XIE Wen, WANG Ruonan, YANG Xin, et al Research on multi-scale residual UNet fused with depthwise separable convolution in PolSAR terrain classification[J]. Journal of Electronics and Information Technology, 2023, 45 (8): 2975- 2985 doi: 10.11999/JEIT220867

[15]	GUAN H L, REN Y, TANG H H, et al Intelligent fault diagnosis methods for hydraulic components based on information fusion: review and prospects[J]. Measurement Science and Technology, 2024, 35 (8): 082001 doi: 10.1088/1361-6501/ad437e

[16]	BI C, YUE X K, DANG Z H, et al A dynamic fault diagnosis method for gravitational reference sensor based on informer-RS[J]. IEEE Sensors Journal, 2025, 25 (2): 3982- 3997 doi: 10.1109/JSEN.2024.3510739

[17]	LEE S, KIM Y, CHOI H J, et al Uncertainty-aware fault diagnosis of rotating compressors using dual-graph attention networks[J]. Machines, 2025, 13 (8): 673 doi: 10.3390/machines13080673

[18]	ZHU X C, RUAN Q S, QIAN S, et al A hybrid model based on transformer and mamba for enhanced sequence modeling[J]. Scientific Reports, 2025, 15 (1): 11428 doi: 10.1038/s41598-025-87574-8

[1]	Minghua ZHAO,Yuxuan LYU,Jiahao LYU,Yifei CHEN,Cheng SHI,Jing HU. Video anomaly detection based on multi-scale appearance and motion fusion[J]. Journal of ZheJiang University (Engineering Science), 2026, 60(8): 1730-1738.

Viewed

Full text

Abstract

Cited

Shared

Discussed