基于光纤光栅传感器的精冲机换向阀阀芯卡滞间接故障诊断

doi:10.3969/j.issn.1004-132X.2026.09.004

摘要/Abstract

摘要：

液压式精冲机作为高端制造装备，其液压阀的健康与否关乎其寿命与加工精度。目前对阀芯卡滞故障诊断研究主要采用人工搭建的平台开展试验，未在工业场景得到验证。为此，提出了一种基于光纤布拉格光栅（FBG）传感器与一维卷积神经网络（1DCNN）的精冲机快进系统换向阀阀芯卡滞故障诊断方法。以HFB-500精冲机为对象，通过AMESim仿真研究阀芯卡滞对滑块与阀的影响进而确定FBG传感器的测点与数据采集方法，而后将试验数据通过Savitzky-Golay（SG）滤波以及去除基线漂移处理，结果表明FBG传感器能够准确捕获阀芯卡滞故障下快进时间会增加这一特征。最后利用1DCNN对阀芯卡滞故障分类，准确率达99.9%，通过与支持向量机（SVM）及长短时记忆网络（LSTM）分类性能的对比验证了1DCNN的优异性能，能够准确识别阀芯卡滞故障。

关键词: 光纤光栅传感器, 精冲机, 换向阀, 阀芯卡滞

Abstract:

As high-end manufacturing equipment， the fine blanking press was directly affected by the health of its hydraulic valves in terms of service life and machining accuracy. Current research on diagnosing spool sticking faults has primarily relied on manually constructed testing platforms， lacking validation in industrial environments. Therefore， a fault diagnosis method was proposed for spool sticking in the rapid approach system directional control valve of the fine blanking press， based on FBG sensor and 1D convolutional neural network （1DCNN）. Using the HFB-500 fine blanking press as the test subject， AMESim simulations were conducted to investigate the impact of spool sticking on the slide and valve， thereby determining the measurement points and data acquisition methods for the FBG sensor. Subsequently， experimental data were processed through Savitzky-Golay（SG） filtering and baseline drift removal. The results demonstrate that the FBG sensor may accurately capture the characteristic increase in rapid approach time during spool sticking faults. Finally， a classification accuracy of 99.9% was achieved by 1DCNN for spool sticking faults. The superior performance of 1DCNN was verified through comparative analyses with support vector machine（SVM） and long short-term memory（LSTM） models， demonstrating its capability to accurately identify spool sticking faults.

Key words: fiber Bragg grating（FBG） sensor, fine blanking press, directional valve, spool sticking

中图分类号:

TG385

魏文婷, 张书照, 刘艳雄, 华林. 基于光纤光栅传感器的精冲机换向阀阀芯卡滞间接故障诊断[J]. 中国机械工程, 2026, 37(9): 2115-2123.

WEI Wenting, ZHANG Shuzhao, LIU Yanxiong, HUA Lin. Indirect Fault Diagnosis of Spool Sticking in Fine Blanking Press Directional Control Valve Based on FBG Sensor[J]. China Mechanical Engineering, 2026, 37(9): 2115-2123.

导出引用管理器 EndNote|Ris|BibTeX

链接本文: https://www.cmemo.org.cn/CN/10.3969/j.issn.1004-132X.2026.09.004

https://www.cmemo.org.cn/CN/Y2026/V37/I9/2115

图/表 20

图1 HFB-500液压式精冲机的液压系统

Fig.1 Hydraulic system of HFB-500 hydraulic fine blanking press

图2 快进系统三位四通电磁换向阀

Fig.2 Three-position four-way solenoid directional valve for rapid traverse system

图3 数据采集系统

Fig.3 Data acquisition system

图4 SG滤波参数选择与去噪时域图

Fig.4 SG filter parameter selection and denoisingtime-domain plot

图5 SG滤波与去除基线漂移数据的时域图

Fig.5 Time-domain waveforms of the signal with SGfiltering and baseline correction

图6 1DCNN结构图

Fig.6 Architecture diagram of 1DCNN

表1 1DCNN网络结构与参数

Tab.1 Architecture and parameters of 1DCNN

网络层	参数
卷积层C1	卷积核个数为32，卷积核尺寸为3，步长为1，激活函数为ReLU
最大池化层M1	池化窗口大小为2
卷积层C2	卷积核个数为64，卷积核尺寸为3，步长为1，激活函数为ReLU
最大池化层M2	池化窗口大小为2
卷积层C3	卷积核个数为128，卷积核尺寸为3，步长为1，激活函数为ReLU
最大池化层M3	池化窗口大小为2
展平层F1	将池化层多维的输出向量转换为一维的向量
全连接层D1	神经元个数为128，激活函数为ReLU
全连接层D2	神经元个数为3，激活函数为softmax

图7 单冲程滑块位移仿真曲线及精冲阶段划分

Fig.7 Simulation curve of single-stroke slide displacement and fine blanking phase division

图8 不同卡滞状态对滑块的影响

Fig.8 Effect of different sticking conditions on slider

表2 仿真得到的不同阀芯卡滞状态下各精冲阶段的时间 (s)

Tab.2 Time of each fine blanking phase under different spool sticking conditions obtained from simulation

阀芯卡滞状态	快进时间	检测和精冲时间	快退时间
正常	1.1	1.1	1.3
轻微卡滞	1.2	1.1	1.3
严重卡滞	1.3	1.1	1.3

图9 不同卡滞状态对阀本身的影响

Fig.9 Effect of different spool sticking conditions on valve

图10 FBG传感器采集的单冲程时域图

Fig.10 Single-stroke time-domain diagram acquired by FBG sensor

表3 FBG传感器测得的不同阀芯卡滞状态下各精冲阶段的时间 (s)

Tab.3 Time of each fine blanking phase under different spool sticking conditions obtained from FBG sensor

阀芯卡滞状态	快进时间	检测与精冲时间	快退时间
正常	1.2	1.2	1.4
轻微卡滞	1.3	1.2	1.4
严重卡滞	1.4	1.2	1.4

表4 FBG传感器与仿真结果各阶段误差 (%)

Tab.4 Errors between FBG sensor measurements and simulation results at various stages

阀芯卡滞状态	快进误差	检测与精冲误差	快退误差
正常	8.33	8.33	7.14
轻微卡滞	7.69	8.33	7.14
严重卡滞	7.14	8.33	7.14

表5 样本与标签的设定

Tab.5 Sample and label counts

故障类型	训练样本数	测试样本数	标签
正常	4000	1000	0
轻微卡滞	4000	1000	1
严重卡滞	4000	1000	2

图11 1DCNN最后一折训练过程中的损失与准确率曲线

Fig.11 Training loss and accuracy curves of 1DCNN in the final fold

图12 5折交叉验证下各模型的平均准确率与标准差

Fig.12 Average accuracy and standard deviation of each model under 5-fold cross-validation

图13 测试的混淆矩阵

Fig.13 Test confusion matrix

图14 各模型的分类指标

Fig.14 Classification metrics for each model

图15 各模型的t-SNE聚类可视化结果

Fig.15 The t-SNE clustering visualization results of each model

参考文献 21

[1]	华林，刘艳雄. 复合精冲成形理论与技术［M］. 北京：科学出版社， 2022.
	HUA Lin， LIU Yanxiong. Hybrid Fine Blanking Theory and Technology［M］. Beijing： Science Press， 2022.
[2]	曹连民，郭徽，王庭明，等. 液压支架立柱液压系统阀芯卡滞诊断［J］. 山东科技大学学报（自然科学版）， 2018， 37（5）： 104-111.
	CAO Lianmin， GUO Hui， WANG Tingming， et al. Valve Core Stagnation Diagnosis of Hydraulic System in Hydraulic Support Column［J］. Journal of Shandong University of Science and Technology（Natural Science）， 2018， 37（5）： 104-111.
[3]	刘艳雄，张昌邦，徐志成，等. 液压精冲机工况能耗智能云端监测平台研究［J］. 塑性工程学报， 2022， 29（11）： 25-31.
	LIU Yanxiong， ZHANG Changbang， XU Zhicheng， et al. Research on Intelligent Cloud Working Condition and Energy Consumption Monitoring Platform for Hydraulic Fine Blanking Machine［J］. Journal of Plasticity Engineering， 2022， 29（11）： 25-31.
[4]	WANG Pei， ZHANG Yuxin， LINJAMA M， et al. Fault Identification and Localization for the Fast Switching Valve in the Equal-coded Digital Hydraulic System Based on Hybrid CNN-LSTM Model［J］. Mechanical Systems and Signal Processing， 2025， 224： 112201.
[5]	YIN Hao， XU He， FAN Weiwang， et al. Fault Diagnosis of Pressure Relief Valve Based on Improved Deep Residual Shrinking Network［J］. Measurement， 2024， 224： 113752.
[6]	ZHANG Guangcheng， ZHANG Kai， SHEN Wei， 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.
[7]	SHI Chong， REN Yan， TANG Hesheng， et al. A Fault Diagnosis Method for an Electro-Hydraulic Directional Valve Based on Intrinsic Mode Functions and Weighted Densely Connected Convolutional Networks［J］. Measurement Science and Technology， 2021， 32（8）： 084015.
[8]	MA Dong， LIU Zhihao， GAO Qinhe， et al. Fault Diagnosis of Multi-step Electromagnetic Hydraulic Valve Group Based on Localized Current Signal CS-SVM［J］. Measurement， 2025， 245： 116632.
[9]	VENKATA S K， RAO S. Fault Detection of a Flow Control Valve Using Vibration Analysis and Support Vector Machine［J］. Electronics， 2019， 8（10）： 1062.
[10]	ZHONG Qi， XU Enguang， SHI Yan， 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.
[11]	SHI Jinchuan， YI Jiyan， REN Yan， et al. Fault Diagnosis in a Hydraulic Directional Valve Using a Two-stage Multi-sensor Information Fusion［J］. Measurement， 2021， 179： 109460.
[12]	SHI Jinchuan， REN Yan， TANG Hesheng， 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.
[13]	张逸群，张钰民，娄小平，等. 光纤光栅在飞机结构载荷监测中的发展现状及展望［J］. 电子测量与仪器学报， 2025， 39（9）： 1-15.
	ZHANG Yiqun， ZHANG Yumin， LOU Xiaoping， et al. Research Status and the Prospect of Fiber Bragg Grating in Aircraft Structural Load Monitoring［J］. Journal of Electronic Measurement and Instrumentation， 2025， 39（9）： 1-15.
[14]	田童，李建乐，邓德双，等. 飞行器结构健康监测技术研究进展［J］. 航空制造技术， 2024， 67（13）： 41-67.
	TIAN Tong， LI Jianle， DENG Deshuang， et al. Research Progress of Structural Health Monitoring Technology for Aircraft［J］. Aeronautical Manufacturing Technology， 2024， 67（13）： 41-67.
[15]	ALHUSSEIN A N D， QAID M R T M， AGLIULLIN T， et al. Fiber Bragg Grating Sensors： Design， Applications， and Comparison with Other Sensing Technologies［J］. Sensors， 2025， 25（7）： 2289.
[16]	魏鹏，王泽林，尹渝，等. 基于光纤传感的装甲车滚动轴承故障诊断技术［J］. 兵器装备工程学报， 2023， 44（1）： 175-182.
	WEI Peng， WANG Zelin， YIN Yu， et al. Fault Diagnosis Technology of the Rolling Bearing in an Armored Vehicle Based on Optical Fiber Sensors［J］. Journal of Ordnance Equipment Engineering， 2023， 44（1）： 175-182.
[17]	曾宪旺，孙文磊，王宏伟. 光纤传感技术结合SVM-ELM的风力机齿轮箱故障诊断分析［J］. 机械设计与制造， 2024（4）： 149-153.
	ZENG Xianwang， SUN Wenlei， WANG Hongwei. Fault Diagnosis and Analysis of Wind Turbine Gearbox Based on Optical Fiber Sensing Technology and SVM-ELM［J］. Machinery Design & Manufacture， 2024（4）： 149-153.
[18]	JIANG Hong， WANG Zhichao， CUI Lina， et al. CEEMDAN-MRAL Transformer Vibration Signal Fault Diagnosis Method Based on FBG［J］. Photonics， 2025， 12（5）： 468.
[19]	JANG B W. Damage Detection and Evaluation in Composite Laminates Using Path Length between Adjacent Peaks of AE Signal from Fiber Optic Sensor［J］. Composite Structures， 2025， 374： 119729.
[20]	SHU Yuwen， LIU Yanxiong， XU Zhicheng， et al. Optimization of Hydraulic Fine Blanking Press Control System Based on System Identification［J］. Processes， 2023， 11（1）： 59.
[21]	陈铁，陈一夫，李咸善，等. 基于CEEMDAN-SG-BiLSTM的变压器油中溶解气体体积分数预测［J］. 高压电器， 2023， 59（12）： 168-175.
	CHEN Tie， CHEN Yifu， LI Xianshan， et al. Prediction for Dissolved Gas Concentration in Power Transformer Oil Based on CEEMDAN-SG-BiLSTM［J］. High Voltage Apparatus， 2023， 59（12）： 168-175.