基于注意力增强DNN的航空液压系统直驱电磁阀抗噪故障诊断

崔译文; 王兴坚; 张育玮; 张润芝; 王少萍

doi:10.13700/j.bh.1001-5965.2026.0027

基于注意力增强DNN的航空液压系统直驱电磁阀抗噪故障诊断

doi: 10.13700/j.bh.1001-5965.2026.0027

崔译文¹,
王兴坚^{1, 2},
张育玮^{1, 3, ,},
张润芝¹,
王少萍^{1, 2}

1.
北京航空航天大学自动化科学与电气工程学院，北京 100191
2.
北京航空航天大学宁波创新研究院，宁波 315800
3.
北京航空航天大学飞行器一体化控制全国重点实验室，北京 100191

基金项目:

国家自然科学基金(52275044,62303030,U2233212)

详细信息

通讯作者:
E-mail：zhangyuwei@buaa.edu.cn

中图分类号: V245.1
计量
- 文章访问数: 341
- HTML全文浏览量: 144
- PDF下载量: 8
- 被引次数: 0
出版历程
- 收稿日期: 2026-01-14
- 录用日期: 2026-01-23
- 网络出版日期: 2026-02-02
- 整期出版日期: 2026-08-31

Attention-enhanced DNN for anti-noise fault diagnosis of direct drive solenoid valves in aviation hydraulic systems

CUI Yiwen¹,
WANG Xingjian^{1, 2},
ZHANG Yuwei^{1, 3
, ,},
ZHANG Runzhi¹,
WANG Shaoping^{1, 2}

1.
School of Automation Science and Electrical Engineering，Beihang University，Beijing 100191，China
2.
Ningbo Institute of Technology，Beihang University，Ningbo 315800，China
3.
National Key Laboratory of Aircraft Integrated Flight Control，Beihang University，Beijing 100191，China

Funds:

National Natural Science Foundation of China (52275044,62303030,U2233212)

More Information

Corresponding author: E-mail：zhangyuwei@buaa.edu.cn

摘要

摘要:
直驱电磁阀(DDSV)作为航空液压系统的关键执行元件，其故障诊断面临电磁干扰与噪声耦合的挑战，而传统数据驱动方法鲁棒性不足。为此，提出基于注意力感知权重机制的鲁棒故障诊断方法，通过投影梯度下降(PGD)法生成贴合实际工况的干扰样本，模拟复杂电磁干扰；深度神经网络(DNN)与注意力感知模块(APM)协同训练，动态调整样本权重，提升模型鲁棒性。搭建了航空液压系统DDSV故障诊断验证实验平台，实验结果显示：所提方法可避免极端权重集中，充分利用原始与干扰样本强化抗干扰能力，在干扰条件下的诊断准确率、收敛稳定性等指标均优于其他传统加权策略。
- 直驱电磁阀 /
- 样本加权 /
- 鲁棒性训练 /
- 故障诊断 /
- 电液系统
Abstract:
Direct drive solenoid valves (DDSVs) serve as critical actuators in aviation hydraulic systems, facing challenges in fault diagnosis due to electromagnetic interference and noise coupling, with traditional data-driven methods lacking robustness. In order to solve this issue, this work suggests a reliable defect diagnostic technique based on an attention-aware weighting mechanism: projection gradient descent (PGD) simulates complicated electromagnetic interference by creating perturbation samples that represent actual operating conditions. Deep neural networks (DNN) and an attention perception module (APM) undergo collaborative training, dynamically adjusting sample weights to enhance model robustness. An experimental platform for aviation solenoid valve fault diagnosis validation was established. According to experimental results, this approach outperforms other weighting strategies in metrics like diagnostic accuracy and convergence stability under interference conditions, avoids extreme weight concentration, and fully utilizes both original and perturbed samples to improve interference resistance.
- direct drive solenoid valve /
- sample weighting /
- robust training /
- fault diagnosis /
- electro-hydraulic system

HTML全文

图 1 故障诊断用航空液压系统DDSV测试台结构与组装示意图

Figure 1. Structure and assembly diagram of the aviation hydraulic system DDSV test rig for fault diagnosis

下载: 全尺寸图片幻灯片

图 2 基于注意力的鲁棒感知DNN

Figure 2. Attention-based robust perceptual DNN

下载: 全尺寸图片幻灯片

图 3 数据预处理过程

Figure 3. Data preprocessing procedure

下载: 全尺寸图片幻灯片

图 4 本文方法的混淆矩阵

Figure 4. Confusion matrix of the proposed method

下载: 全尺寸图片幻灯片

图 5 训练损失曲线

Figure 5. Training loss curve

下载: 全尺寸图片幻灯片

图 6 不同加权策略下样本权重的比较

Figure 6. Comparison of sample weights under different weighting strategies

下载: 全尺寸图片幻灯片

参考文献(32)

[1]	Huang K K, Wu S J, Li F B, et al. Fault diagnosis of hydraulic systems based on deep learning model with multirate data samples[J]. IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(11): 6789-6801.
[2]	Shen W, Zhao H M. Fault tolerant control of nonlinear hydraulic systems with prescribed performance constraint[J]. ISA Transactions, 2022, 131: 1-14.
[3]	Ding R Q, Cheng M, Jiang L, et al. Active fault-tolerant control for electro-hydraulic systems with an independent metering valve against valve faults[J]. IEEE Transactions on Industrial Electronics, 2021, 68(8): 7221-7232.
[4]	Pedersen H C, Bak-Jensen T, Jessen R H, et al. Temperature-independent fault detection of solenoid-actuated proportional valve[J]. IEEE/ASME Transactions on Mechatronics, 2022, 27(6): 4497-4506.
[5]	Kong X D, Cai B P, Yu Y L, et al. Intelligent diagnosis method for early faults of electric-hydraulic control system based on residual analysis[J]. Reliability Engineering & System Safety, 2025, 261: 111142.
[6]	Shi J C, Yi J Y, Ren Y, et al. Fault diagnosis in a hydraulic directional valve using a two-stage multi-sensor information fusion[J]. Measurement, 2021, 179: 109460.
[7]	Shen S, Lu H, Sadoughi M, et al. A physics-informed deep learning approach for bearing fault detection[J]. Engineering Applications of Artificial Intelligence, 2021, 103: 104295.
[8]	Ping Z W, Wang D W, Zhang Y, et al. Few-shot aero-engine bearing fault diagnosis with denoising diffusion based data augmentation[J]. Neurocomputing, 2025, 622: 129327.
[9]	Zhang G C, Zhang K, Shen W, et al. Intelligent fault diagnosis of multi-way directional valves in hydraulic systems using digital twin and deep learning approaches[J]. Mechanical Systems and Signal Processing, 2025, 230: 112579.
[10]	Gao Y P, Gao L, Li X Y, et al. A hierarchical training-convolutional neural network for imbalanced fault diagnosis in complex equipment[J]. IEEE Transactions on Industrial Informatics, 2022, 18(11): 8138-8145.
[11]	Men J K, Zhao C M. An adaptive imbalance modified online broad learning system-based fault diagnosis for imbalanced chemical process data stream[J]. Expert Systems with Applications, 2023, 234: 121159.
[12]	Liu J Q, Zhao N. Improved fault-tolerant method and control strategy based on reverse charging for the power electronic traction transformer[J]. IEEE Transactions on Industrial Electronics, 2018, 65(3): 2672-2682.
[13]	Wang J B, Liu K, Wei D, et al. Multielectrical signal fusion-based method for bearing fault diagnosis in permanent magnet synchronous machines under dynamic conditions[J]. IEEE/ASME Transactions on Mechatronics, 2025, 30(1): 787-802.
[14]	Wang X, Jiang H K, Mu M Z, et al. A dynamic collaborative adversarial domain adaptation network for unsupervised rotating machinery fault diagnosis[J]. Reliability Engineering & System Safety, 2025, 255: 110662.
[15]	Chen J H, Li T, Wang J, et al. Closing the simulation-to-reality gap for fault diagnosis in unknown environment: a Sim2Real knowledge transfer approach with contrastive learning[J]. IEEE/ASME Transactions on Mechatronics, 2026, 31(1): 663-674.
[16]	Lu W C, Duan J D, Cheng L, et al. Electromagnetic interference effect assessment under measuring testability limitation based on physics-informed neural network and Gaussian process regression[J]. IEEE Transactions on Power Electronics, 2024, 39(10): 12413-12423.
[17]	Lv S S, Tao C Z, Zhai Y F, et al. Research on noise reduction of pressure pulsation signal of axial flow pump based on global optimization parrot algorithm improved variational mode decomposition[J]. Physics of Fluids, 2025, 37(7): 077185.
[18]	Stosiak M, Karpenko M, Ivannikova V, et al. The impact of mechanical vibrations on pressure pulsation, considering the nonlinearity of the hydraulic valve[J]. Journal of Low Frequency Noise, Vibration and Active Control, 2025, 44(2): 706-719.
[19]	Liu X X, Yin Y B, Zheng S W, et al. The modeling and analysis of operating characteristics and temperature drift of deflector jet servo valve under the wide temperature range[J]. Proceedings of the Institution of Mechanical Engineers, Part C: Journal of Mechanical Engineering Science, 2024, 238(12): 5921-5936.
[20]	Xu Y P, Guo K, Li J F, et al. A novel rotational actuator with variable stiffness using S-shaped springs[J]. IEEE/ASME Transactions on Mechatronics, 2021, 26(4): 2249-2260.
[21]	Zhong Q, Xu E G, Shi Y, et al. Fault diagnosis of the hydraulic valve using a novel semi-supervised learning method based on multi-sensor information fusion[J]. Mechanical Systems and Signal Processing, 2023, 189: 110093.
[22]	Li K J, Xie Y Z, Chen Y H, et al. Bayesian inference for susceptibility of electronics to transient electromagnetic disturbances with failure mechanism consideration[J]. IEEE Transactions on Electromagnetic Compatibility, 2020, 62(5): 1669-1677.
[23]	Song C, Alkhalifah T A. Wavefield reconstruction inversion via physics-informed neural networks[J]. IEEE Transactions on Geoscience and Remote Sensing, 2022, 60: 5908012.
[24]	Qu B G, Wang Z D, Shen B, et al. Adaptive decentralized state estimation for multimachine power grids under measurement noises with unknown statistics[J]. IEEE Transactions on Industrial Informatics, 2025, 21(2): 1655-1664.
[25]	Dong Z L, Jiang Y H, Jiao W D, et al. Double attention-guided tree-inspired grade decision network: a method for bearing fault diagnosis of unbalanced samples under strong noise conditions[J]. Advanced Engineering Informatics, 2025, 64: 103004.
[26]	Moosavi-Dezfooli S M, Fawzi A, Frossard P. DeepFool: a simple and accurate method to fool deep neural networks[EB/OL]. (2016-07-04)[2026-01-10]. https://arxiv.org/abs/1511.04599.
[27]	Castejón-Limas M, Alaiz-Moreton H, Fernández-Robles L, et al. Robust weighted regression via PAELLA sample weights[J]. Neurocomputing, 2020, 391: 325-333.
[28]	Yang X Q, Liu X, Yin S. Robust identification of nonlinear systems with missing observations: the case of state-space model structure[J]. IEEE Transactions on Industrial Informatics, 2019, 15(5): 2763-2774.
[29]	Chen Y H, Xie Y Z, Ge X Y, et al. Vulnerability assessment of equipment excited by disturbances based on support vector machine and Gaussian process regression[J]. IEEE Transactions on Electromagnetic Compatibility, 2021, 63(1): 103-110.
[30]	Gao R Z, Liu F, Zhou K W, et al. Local reweighting for adversarial training[EB/OL]. (2021-06-30)[2026-01-10]. https://arxiv.org/abs/2106.15776.
[31]	Zhang J F, Zhu J N, Niu G, et al. Geometry-aware instance-reweighted adversarial training[EB/OL]. (2021-05-31)[2026-01-10]. https://arxiv.org/abs/2010.01736.
[32]	Fu M J, Wang X M, Wang J, et al. Prototype Bayesian meta-learning for few-shot image classification[J]. IEEE Transactions on Neural Networks and Learning Systems, 2025, 36(4): 7010-7024.