基于改进CNN-LSTM的开关闩故障诊断方法

doi:10.12382/bgxb.2024.0818

摘要/Abstract

摘要：

针对某舰炮开关闩机构关重件磨损和弹簧弹性减弱两类典型故障模式,传统故障诊断方法主要依赖于人工检查、专家经验推理和理论仿真等方法,不仅时间周期较长,而且诊断精度难以保证。针对此问题,采用深度学习的方法,提出一种基于麻雀搜索算法(Sparrow Search Algorithm,SSA)的格拉姆角场结合卷积神经网络和长短期记忆神经网络(Graham Angle Field-Convolutional Neural Network-Long Short-Term Memory,GAF-CNN-LSTM)的开关闩故障诊断方法。通过试验台架采集开关闩机构故障原始信号并进行预处理,通过时频分析法和格拉姆角场法建立一维时序数据和二维图像故障数据集;将故障数据集分别输入到使用卷积神经网络(Convolutional Neural Network,CNN)和长短期记忆神经网络(Long Short-Term Memory,LSTM)通道中,利用CNN强大的空间特征提取能力和LSTM挖掘数据时序特征能力进行特征提取,并将二者得到的特征信息进行融合,在全连接层和激活函数的作用下输出诊断结果;通过SSA对GAF-CNN-LSTM网络结构中的超参数进行优化,提高模型的诊断精度和适用性。经测试数据验证:提出的SSA-GAF-CNN-LSTM故障诊断模型不仅可以更精准地诊断开关闩机构故障类型,而且具有更强的泛化能力及抗干扰能力,有效地提高了开关闩机构故障诊断的性能。

关键词: 故障诊断, 开关闩机构, 深度学习算法, 群体智能优化算法, 格拉姆角场

Abstract:

In view of the two typical failure modes of breechblock opening-closing mechanism for a naval gun,namely the wear of the key parts and the weakening of spring elasticity,the traditional fault diagnosis methods mainly rely on manual inspection,expert empirical reasoning and theoretical simulation.However,these methods not only take a long time for diagnosis,but also the diagnostic accuracy cannot be guarantee.In order to solve this problem,a fault diagnosis method of Gram angle field combined with convolutional neural network and long short-term memory neural network (GAF-CNN-LSTM) based on sparrow search algorithm (SSA) is proposed by using the deep learning algorithms.Firstly,the original fault signal of the breechblock opening-closing mechanism is collected and preprocessed by the test bench,and the one-dimensional time-series data and two-dimensional image fault dataset are established by the time-frequency analysis method and the Gramian angular field method.Then the fault dataset is input into the LSTM and CNN channels,respectively,and the powerful spatial feature extraction ability of CNN and the time-series feature ability of LSTM mining data are used to extract the features,and the feature informations obtained by the two abilities are fused to output the diagnostic results under the action of the fully connected layer and activation function.Finally,the SSA optimization algorithm is used to optimize the hyperparameters in the GAF-CNN-LSTM network structure to improve the diagnostic accuracy and applicability of the model.The proposed SSA-GAF-CNN-LSTM fault diagnosis model is verified by the test data.The result shows that the proposed fault diagnosis model can not only diagnose the fault type of the breechblock opening-closing mechanism for naval gun more accurately,but also has stronger generalization ability and anti-interference ability,which effectively improves the fault diagnosis performance of the breechblock opening-closing mechanism.

Key words: fault diagnosis, breechblock opening-closing mechanism, deep learning algorithm, swarm intelligence optimization algorithm, Gramian angular field

高振华, 秦奋起, 王琳琳, 于存贵. 基于改进CNN-LSTM的开关闩故障诊断方法[J]. 兵工学报, 2025, 46(9): 240818-.

GAO Zhenhua, QIN Fenqi, WANG Linlin, YU Cungui. An Improved CNN-LSTM-based Fault Diagnosis Method for Breechblock Opening-closing Mechanism[J]. Acta Armamentarii, 2025, 46(9): 240818-.

图/表 17

图1 炮尾振动加速度信号

Fig.1 Breech vibration acceleration signal

图2 关闩弹簧弹性减弱EMD结果

Fig.2 EMD results of weakening of latch spring elasticity

表1 关闩簧弹性减弱IMF分量相关系数

Table 1 The elasticity of the latch spring weakens the correlation coefficient of the IMF component

IMF 分量	IMF1	IMF2	IMF3	IMF4	IMF5	IMF6
相关系数	0.95919	0.27252	0.178269	0.104951	0.059511	0.047242

图3 闩体速度信号

Fig.3 Breechblock velocity signal

表2 不同故障工况GASF图像

Table 2 GASF images for different fault cases

故障类型	GASF图像
正常状态
开闩弹簧弹性减弱
曲臂磨损

图4 传统CNN-LSTM模型结构

Fig.4 Structure of the traditional CNN-LSTM model

图5 GAF-CNN-LSTM模型结构

Fig.5 Structure of CNN-LSTM model

图6 训练模型迭代

Fig.6 Iterative curves for training the model

表3 开关闩机构典型故障模式标签

Table 3 Typical failure mode labels for breechblock opening-closing mechanism

故障模式	故障标签
抽筒子、闩体挂臂磨损	1
滚轮磨损	2
关闩簧弹性减弱	3
开闩模板磨损	4
开闩弹簧弹性减弱	5
曲臂磨损	6
起落弹簧弹性减弱	7

表4 模型主要参数

Table 4 The main parameters of the model

故障特征类别	网络层	节点大小	步长	卷积核数量
一维	卷积层1	2×1	1×1	32
	池化层1	2×1	2×1
	卷积层2	2×1	1×1	16
	LSTM层	60
	全连接层1	10×1		1
	全连接层2	7×1		1
二维	卷积层1	2×2×3	1×1	32
	池化层1	2×2	2×2
	卷积层2	2×2×3	1×1	16

表5 优化后超参数取值

Table 5 The values of the optimized hyperparameters

算法	学习率	卷积层1 大小	卷积核1 数量	卷积层2 大小	卷积核2 数量	LSTM神经元个数
GWO 算法	0.00717	2	23	2	10	6
SSA	0.001	5	21	2	19	51
WOA	0.00854	4	11	5	26	95

图7 不同优化算法寻优迭代曲线

Fig.7 Optimization iteration curves for different optimization algorithms

图8 故障诊断流程

Fig.8 The process of troubleshooting

图9 试验装置

Fig.9 Test setup

图10 实验仪器布局

Fig.10 Experimental instrument layout

表6 诊断结果对比

Table 6 Comparison of diagnostic results %

诊断模型	最高准确率	最低准确率	10次平均准确率
CNN	88.09	80.09	85.71
LSTM	90.47	80.09	86.42
CNN-LSTM	95.23	85.71	90.23
GAF-CNN-LSTM	100.00	88.09	96.19
GWO-GAF-CNN-LSTM	100.00	95.23	98.09
WOA-GAF-CNN-LSTM	100.00	97.61	99.04
SSA-GAF-CNN-LSTM	100.00	97.61	99.28

图11 不同诊断模型结果对比

Fig.11 Comparison of diagnostic results from different diagnostic models

参考文献 25

[1]	付帅. 某火炮开闩动力学特性分析及可靠性研究[D]. 南京: 南京理工大学, 2014.
	FU S. Analysis of dynamic characteristics and reliability of a Breechblock-opening of gun[D]. Nanjing: Nanjing University of Science and Technology, 2014. (in Chinese)
[2]	付彩越, 王长庚. 基于故障树的某舰炮提前关闩故障分析[J]. 兵工自动化, 2018, 37(8):67-72.
	FU C Y, WANG C G. Failure analysis of closing breech ahead of time based on fault tree for certain type naval gun[J]. Ordnance Industry Automation, 2018, 37(8):67-72. (in Chinese)
[3]	付昆. 某火炮炮闩系统故障预测研究及关键件寿命分析[D]. 南京: 南京理工大学, 2012.
	FU K. Research on fault prediction of a gun breech system and life analysis of key parts[D]. Nanjing: Nanjing University of Science and Technology, 2012. (in Chinese)
[4]	胡慧斌, 侯小锋, 曹立军, 等. 击发装置关重件磨损特性分析及失效寿命预测[J]. 火炮发射与控制学报, 2014, 35(4):62-66.
	HU H B, HOU X F, CAO L J, et al. Wearing characteristics analysis and failure life forecast of key components in gun firing mechanism[J]. Journal of Gun Launch & Control, 2014, 35(4):62-66. (in Chinese)
[5]	YANG X F, ZHANG F P, WEI J F. Fuzzy RBF neural network fault diagnosis method based on knowledge and data fusion for the recoil system[J]. Journal of Intelligent & Fuzzy Systems, 2024, 46(2):4981-4994.
[6]	李英顺, 于昂, 李茂, 等. 基于KLDA-IDBO-BP的装甲车发动机故障诊断[J]. 兵工学报, 2025, 46(3):240083. doi: 10.12382/bgxb.2024.0083
	LI Y S, YU A, LI M, et al. Armored vehicle engine fault diagnosis based on KLDA-IDBO-BP[J]. Acta Armamentarii, 2025, 46(3):240083. (in Chinese)
[7]	姜斌, 程月华, 孙颢, 等. 一种利用LSTM-FCN的导弹舵回路故障诊断算法[J]. 宇航学报, 2023, 44(5):687-698.
	JIANG B, CHENG Y H, SUN H, et al. A fault diagnosis algorithm of missile rudder loop using LSTM-FCN[J]. Journal of Astronautics, 2023, 44(5):687-698. (in Chinese)
[8]	TONG S G, FU Z L, TONG Z M, et al. Fault diagnosis for gearboxes based on Fourier decomposition method and resonance demodulation[J]. Journal of Zhejiang University-Science A(Applied Physics & Engineering), 2023, 24(5):404-419.
[9]	LI C Y, CUI D X. Fault diagnosis of marine diesel engine based on multi-scale time domain decomposition and convolutional neural network[J]. Polish Maritime Research, 2024, 31(3):85-93.
[10]	张学飞. 某大口径火炮装填装置液压系统故障诊断研究[D]. 南京: 南京理工大学, 2021.
	ZHANG X F. Research on the fault diagnosis of the hydraulic system of a large-caliber artillery loading device[D]. Nanjing: Nanjing University of Science and Technology, 2021. (in Chinese)
[11]	杨宁. 某大口径火炮药协调装置故障诊断研究[D]. 南京: 南京理工大学, 2021.
	YANG N. Research on fault diagnosis of a large-caliber artillery ammunition coordination device[D]. Nanjing: Nanjing University of Science and Technology, 2021. (in Chinese)
[12]	邵夕安. 火炮供输弹系统健康监测技术研究[D]. 南京: 南京理工大学, 2019.
	SHAO X A. Research on health monitoring technology of artillery ammunition supply and delivery system[D]. Nanjing: Nanjing University of Science and Technology, 2019. (in Chinese)
[13]	黄文宽, 钱林方, 尹强, 等. 基于迁移学习的供药装置故障诊断方法[J]. 兵工学报, 2023, 44(10):2964-2974. doi: 10.12382/bgxb.2022.0767
	HUANG W K, QIAN L F, YIN Q, et al. Fault diagnosis method of modular charge feeding mechanism based on transfer learning[J]. Acta Armamentarii, 2023, 44(10):2964-2974. (in Chinese) doi: 10.12382/bgxb.2022.0767
[14]	LI C Y, HU Y H, JIANG J W, et al. Fault diagnosis of a marine power-generation diesel engine based on the Gramian angular field and a convolutional neural network[J]. Journal of Zhejiang University-Science A(Applied Physics & Engineering), 2024, 25(6):470-482. (in Chinese)
[15]	崔江, 周凡, 陈永凡, 等. 一种基于GAMF-CNN的航空发电机整流器故障诊断技术[J]. 航空学报, 2024, 45(24):237-248.
	CUI J, ZHOU F, CHEN Y F, et al. A technique of aerospace generator rectifier fault diagnosis based on GAMF-CNN[J]. Acta Aeronautica et Astronautica Sinica, 2024, 45(24):237-248. (in Chinese)
[16]	胡小利, 孙苗苗, 徐宁骏. 舰炮武器装备智能化维修保障研究[J]. 兵器装备工程学报, 2023, 44(7):135-141.
	HU X L, SUN M M, XU N J. Study on intelligent maintenance and support of naval gun weapons[J]. Journal of Ordnance Equipment engineering, 2023, 44(7):135-141. (in Chinese)
[17]	代普, 李永锋, 韩建刚, 等. 基于CAN总线和大容量数据存储的舰炮远程监控系统设计[J]. 兵工学报, 2022, 43(增刊1):88-96.
	DAI P, LI Y F, HAN J G, et al. Design of naval gun remote monitoring system based on CAN bus and large capacity data storage[J]. Acta Armamentarii, 2022, 43(S1):88-96. (in Chinese) doi: 10.12382/bgxb.2022.A006
[18]	HUANG N E, SHEN Z, LONG S R, et al. The empirical mode decomposition and the Hilbert spectrum for nonlinear and nonstationary time series analysis[J]. Proceedings of the Royal Society A:Mathematica, Physical and Engineering Sciences,1998, 454(1971):903-995.
[19]	唐伯宇, 邵星, 王翠香, 等. 基于双通道CNN与SSA-SVM的滚动轴承故障诊断[J]. 控制理论与应用, 2024, 41(9):1626-1635.
	TANG B Y, SHAO X, WANG C X, et al. Rolling bearing fault diagnosis based on dual channel CNN and SSA-SVM[J]. Control Theory & Applications, 2024, 41(9):1626-1635. (in Chinese)
[20]	赵妍, 唐文石, 聂永辉, 等. 基于格拉姆角差场和卷积神经网络的宽频振荡分类方法[J]. 电网技术, 2022, 46(11):4364-4372.
	ZHAO Y, TANG W S, NIE Y H, et al. Broadband oscillation classification method based on GADF-AlexNet[J]. Power System Technology, 2022, 46(11):4364-4372. (in Chinese)
[21]	姚洋, 陈涛, 贾然, 等. 基于改进CNN-LSTM的综合传动装置异常检测方法[J]. 机电工程, 2024, 41(11):1986-1994.
	YAO Y, CHEN T, JIA R, et al. Detection method of integrated transmission system anomaly based on improved CNN-LSTM[J]. Journal of Mechanical & Electrical Engineering, 2024, 41(11):1986-1994. (in Chinese)
[22]	BENI G, WANG J. Swarm intelligence in cellular robotic systems[J]. Robots and biological systems:towards a new bionics, 1993, 102:703-712.
[23]	MIRJALILI S, MIRJALILI S M, LEWIS A. Grey wolf optimizer[J]. Advances in Engineering Software, 2014, 69 (3):46-61.
[24]	XUE J K, SHEN B. A novel swarm intelligence optimization approach:sparrow search algorithm[J]. Systems Science & Control Engineering, 2020, 8(1):22-34.
[25]	MIRJALILI S, LEWIS A. The whale optimization algorithm[J]. Advances in Engineering Software, 2016, 95:51-67.