作者：李从跃¹, 胡以怀¹, 沈威¹, 崔德馨¹, 张成², 芮晓松²

作者单位：1. 上海海事大学商船学院，上海 201306;
2. 招商局鼎衡造船有限公司，江苏 扬州 225217

关键词：连续小波变换;卷积神经网络;小波时频图;船用机械;故障诊断

摘要：

针对船用机械故障特征自适应提取与智能化诊断问题，采用连续小波变换与卷积神经网络的船舶机械故障诊断方法。以船用风机为例，首先模拟船用机械不同故障并采集振动信号，通过连续小波变换将一维振动信号转化为特征图谱，其包含大量的时频信息。然后通过多次训练后，确定网络结构参数，建立卷积神经网络结构，将时频图作为卷积神经网络输入，挖掘更深层次的高度抽象的故障特征信息。最后在卷积神经网络的输出层接入softmax分类器，实现船用机械的故障诊断。实验结果表明：所提方法能准确识别故障类型，且具有较强的鲁棒性和泛化能力，诊断准确率可达99.3%。与集成经验模态分解、极限学习机故障诊断方法相比，该方法有更高的诊断精度。

Fault diagnosis of marine machinery based on wavelet transform and CNN
LI Congyue¹, HU Yihuai¹, SHEN Wei¹, CUI Dexin¹, ZHANG Cheng², RUI Xiaosong²
1. Merchant Marine College, Shanghai Maritime University, Shanghai 201306, China;
2. China Merchants Dingheng Shipbuilding Co., Ltd., Yangzhou 225217, China
Abstract: Aiming at the problem of self-adaptive extraction and intelligent diagnosis of marine machinery fault features, a marine machinery fault diagnosis method using continuous wavelet transform and convolutional neural network is proposed. Taking marine wind turbines as an example, firstly, different faults of marine machinery are simulated and vibration signals are collected, and the one-dimensional vibration signals are converted into feature maps by continuous wavelet transform, which contain a large amount of time-frequency information. Then after several trainings, the network structure parameters are determined, the convolutional neural network structure is established, and the time-frequency graph is used as the input of the convolutional neural network to mine deeper and highly abstract fault feature information. Finally, the softmax classifier is connected to the output layer of the convolutional neural network to realize the fault diagnosis of marine machinery. The experimental results show that the proposed method can accurately identify fault types and has strong robustness and generalization ability, and the diagnostic accuracy can reach 99.3%. Compared with integrated empirical mode decomposition and extreme learning machine fault diagnosis methods, this method has higher diagnostic accuracy.
Keywords: continuous wavelet transform;convolutional neural network;wavelet time-frequency map;marine machinery;fault diagnosis
2024, 50(3):183-192  收稿日期: 2022-04-28;收到修改稿日期: 2022-06-29
基金项目: 上海市科技计划（20DZ2252300）
作者简介: 李从跃(1998-),男,山东德州市人,硕士研究生,专业方向为船舶动力装置智能故障诊断。
参考文献
[1] 但长林, 李三雁, 张彬. 基于样本熵和SVM的滚动轴承故障诊断方法研究[J]. 中国测试, 2020, 46(11): 37-42
DAM C L, LI S Y, ZHANG B. Research on fault diagnosis method for rolling element bearings based on sample entropy and SVM[J]. China Measurement & Test, 2020, 46(11): 37-42
[2] 蒋佳炜, 胡以怀, 柯赟, 等. 基于小波包特征提取和模糊熵特征选择的柴油机故障分析[J]. 振动与冲击, 2020, 39(4): 273-277, 298
JIANG J W, HU Y H, KE Y, et al. Fault diagnosis of diesel engines based on wavelet packet energy spectrum feature extraction and fuzzy entropy feature selection[J]. Journal of Vibration and Shock, 2020, 39(4): 273-277, 298
[3] 蔡波, 黄晋英, 杜金波, 等. 基于MEEMD多特征融合与LS-SVM的行星齿轮箱故障诊断[J]. 中国测试, 2021, 47(9): 126-132
CAI B, HUANG J Y, DU J B, et al. Planetary gearbox fault diagnosis based on MEEMD multi-feature fusion and LS-SVM[J]. China Measurement & Test, 2021, 47(9): 126-132
[4] 汪祖民, 张志豪, 秦静, 等. 基于卷积神经网络的机械故障诊断技术综述[J]. 计算机应用, 2022, 42(4): 1036-1043
WANG Z M, ZHANG Z H, QIN J, et al. Review of mechanical fault diagnosis technology based on convolutional neural network[J]. Journal of Computer Applications, 2022, 42(4): 1036-1043
[5] HINTON G E, OSINDERO S, TEH Y W. A fast learning algorithm for deep belief nets.[J]. Neural computation, 2006, 18(7): 1527-1554
[6] 杨劼立, 林建辉, 谌亮. 基于MED辅助特征提取CNN模型的列车轴承故障诊断方法[J]. 中国测试, 2020, 46(10): 124-129
YANG J L, LIN J H, CHEN L. Fault diagnosis method for train bearings based on MED-assisted feature extraction CNN model[J]. China Measurement & Test, 2020, 46(10): 124-129
[7] 赵小强, 张青青. 改进Alexnet的滚动轴承变工况故障诊断方法[J]. 振动、测试与诊断, 2020, 40(3): 472-480, 623
ZHAO X Q, ZHANG Q Q. Improved alexnet based fault diagnosis method for rolling bearing under variable conditions[J]. Journal of Vibration, Measurement & Diagnosis, 2020, 40(3): 472-480, 623
[8] 李俊, 刘永葆, 余又红. 卷积神经网络和峭度在轴承故障诊断中的应用[J]. 航空动力学报, 2019, 34(11): 2423-2431
LI J, LIU Y B, YU Y H. Application of convolutional neural network and kurtosis in fault diagnosis of rolling bearing[J]. Journal of Aerospace Power, 2019, 34(11): 2423-2431
[9] CHEN Z Q, LI C, SANCHEZ R V. Gearbox fault identification and classification with convolutional neural networks[J]. Shock and Vibration, 2015(2): 1-10
[10] 肖雄, 王健翔, 张勇军, 等. 一种用于轴承故障诊断的二维卷积神经网络优化方法[J]. 中国电机工程学报, 2019, 39(15): 4558-4568
XIAO X, WANG J X, ZHANG Y J, et al. A two-dimensional convolutional neural network optimization method for bearing fault diagnosis[J]. Proceedings of the CSEE, 2019, 39(15): 4558-4568
[11] 陈仁祥, 黄鑫, 杨黎霞, 等. 基于卷积神经网络和离散小波变换的滚动轴承故障诊断[J]. 振动工程学报, 2018, 31(5): 883-891
CHEN R X, HUANG X, YANG L X, et al. Rolling bearing fault identification based on convolution neural network and discrete wavelet transform[J]. Journal of Vibration Engineering, 2018, 31(5): 883-891
[12] LECUN L, BOTTOU Y, BENGIO P H. Gradient-based learning applied to document recognition[J]. Proceedings of the IEEE, 1998, 86(11): 2278-2324
[13] 赵小强, 张亚洲. 利用改进卷积神经网络的滚动轴承变工况故障诊断方法[J]. 西安交通大学学报, 2021, 55(12): 108-118
ZHAO X Q, ZHANG Y Z. Improved CNN-based fault diagnosis method for rolling bearings under variable working conditions[J]. Journal of Xi'an Jiaotong University, 2021, 55(12): 108-118
[14] 谢旭阳, 余刃, 王天舒, 等. 基于卷积神经网络和迁移学习的电动泵故障诊断方法研究[J]. 兵器装备工程学报, 2021, 42(7): 239-245
XIE X Y, YU R, WANG T S, et al. Fault diagnosis method for electric pump based on convolutional neural network and transfer learning[J]. Journal of Ordnance Equipment Engineering, 2021, 42(7): 239-245
[15] 吴定海, 任国全, 王怀光, 等. 基于卷积神经网络的机械故障诊断方法综述[J]. 机械强度, 2020, 42(5): 1024-1032
WU D H, REN G Q, WANG H G, et al. The review of mechanical fault diagnosis methods based on convolutional neural network[J]. Journal of Mechanical Strength, 2020, 42(5): 1024-1032
[16] 刘宁, 刘四新, 奚佳欣, 等. 基于S变换的简支梁结构损伤检测方法研究[J]. 振动与冲击, 2018, 37(9): 161-167, 172
LIU N, LIU S X, XI J X, et al. Damage detection method for simply supported RC beams based on S transformation[J]. Journal of Vibration and Shock, 2018, 37(9): 161-167, 172
[17] 王超, 毛羚. 基于VMD和广义Morse小波的结构瞬时频率识别[J]. 振动、测试与诊断, 2020, 40(5): 957-962, 1026
WANG C, MAO L. Shock resistance of vibration isolation system with hydraulic displacement[J]. Journal of Vibration, Measurement & Diagnosis, 2020, 40(5): 957-962, 1026
[18] 褚亚奇, 朱波, 赵新刚, 等. 基于时空特征学习卷积神经网络的运动想象脑电解码方法[J]. 生物医学工程学杂志, 2021, 38(1): 1-9
CHU Y Q, ZHU B, ZHAO X G, et al. Convolutional neural network based on temporal-spatial feature learning for motor imagery electroencephalogram signal decoding[J]. Journal of Biomedical Engineering, 2021, 38(1): 1-9