一种基于多通道马尔可夫变迁场的故障诊断方法

doi:10.13229/j.cnki.jdxbgxb20210669

吉林大学学报(工学版) ›› 2022, Vol. 52 ›› Issue (2): 491-496.doi: 10.13229/j.cnki.jdxbgxb20210669

• 车辆工程·机械工程 • 上一篇

一种基于多通道马尔可夫变迁场的故障诊断方法

曹洁^1,²(),马佳林¹,黄黛麟¹,余萍³()

^1.兰州理工大学计算机与通信学院，兰州 730050
^2.甘肃省制造信息工程研究中心，兰州 730050
^3.兰州理工大学电气工程与信息学院，兰州 730050

收稿日期:2021-07-16 出版日期:2022-02-01 发布日期:2022-02-17
通讯作者: 余萍 E-mail:caoj@lut.edu.cn;yup@lut.edu.cn
作者简介:曹洁（1966-），女，教授，博士生导师.研究方向：信息检测与估计，智能信息处理，机器视觉信息获取与处理，模式识别理论及应用，智能交通系统，计算机控制技术.E-mail：caoj@lut.edu.cn
基金资助:
国家自然科学基金项目(61763028);甘肃省教育厅项目(2021CXZX-517)

A fault diagnosis method based on multi Markov transition field

Jie CAO^1,²(),Jia-lin MA¹,Dai-lin HUANG¹,Ping YU³()

^1.College of Computer and Communication，Lanzhou University of Technology，Lanzhou 730050，China
^2.Engineering Research Center of Manufacturing Information of Gansu Province，Lanzhou 730050，China
^3.College of Electrical & Information Engineering，Lanzhou University of Technology，Lanzhou 730050，China

Received:2021-07-16 Online:2022-02-01 Published:2022-02-17
Contact: Ping YU E-mail:caoj@lut.edu.cn;yup@lut.edu.cn

摘要/Abstract

摘要：

深度学习在故障诊断中有良好的诊断能力与泛化能力，但大部分工作是直接从卷积层面上提取信号特征图，使邻近信号点未被考虑，并且采样频率不同也会对特征提取有影响。为此，本文基于MTF以及ResNet18算法提出了M2TF-ResNet算法。本文在凯斯西储大学（CWRU）轴承数据集中进行了大量实验。通过验证得出：该算法可适应不同采样频率下信号的特征提取，避免训练过拟合，并且与其他故障诊断方式相比，该算法在诊断率上的优势更突出。

关键词: 故障诊断, 深度学习, 马尔可夫变迁场, 残差神经网络

Abstract:

Deep learning has good diagnostic capabilities and generalization capabilities in fault diagnosis， but most of the work is to directly extract signal feature maps from the convolutional layer so that adjacent signal points are not considered， and different sampling frequencies will also affect feature extraction. Therefore， the M2TF-ResNet algorithm was proposed based on the MTF and ResNet18 algorithm. Many experiments were carried out in Case Western Reserve University （CWRU） bearing dataset. Through the verification， it can adapt to the signal feature extraction under different sampling frequencies and avoid over-fitting training. And compared with other fault diagnosis methods， it has more prominent advantages in the diagnosis rate.

Key words: fault diagnosis, deep learning, Markov transition field, residual network

中图分类号:

TP277

曹洁,马佳林,黄黛麟,余萍. 一种基于多通道马尔可夫变迁场的故障诊断方法[J]. 吉林大学学报(工学版), 2022, 52(2): 491-496.

Jie CAO,Jia-lin MA,Dai-lin HUANG,Ping YU. A fault diagnosis method based on multi Markov transition field[J]. Journal of Jilin University(Engineering and Technology Edition), 2022, 52(2): 491-496.

图/表 7

图1

图2

图3

表1

ResNet18的结构"

网络层名称	参数	输出
Conv Block 1	7×7，64， strider 2	112×112
Conv Block 2，x	3×3 max pool， strider 2 $3 × 3, 64 3 × 3, 64 × 2$ dropout 0.2	56×56
Conv Block 3，x	$3 × 3, 128 3 × 3, 128 × 2$ dropout 0.2	28×28
Conv Block 4，x	$3 × 3, 256 3 × 3, 256 × 2$ dropout 0.2	14×14
Conv Block 5，x	$3 × 3, 512 3 × 3, 512 × 2$ dropout 0.2	7×7
Classification	Average pool， fc， SoftMax	1×4

表1

表2

CWRU轴承数据设定"

数据集名称	电机转速/（r $?$ min $- 1$ ）	电机载荷	采样点数目	训练集：测试集
CWRU 1797	1797	0	400	4 $∶$ 1
CWRU 1772	1772	1	400
CWRU 1750	1750	2	400
CWRU 1730	1730	3	400

表2

表3

图4

参考文献 18

1	Ghate V N, Dudul S V. Design of optimal MLP and RBF neural network classifier for fault diagnosis of three phase induction motor[J]. International Journal of Advanced Mechatronic Systems, 2010, 2(3): 204-216.
2	Keleolu C, KüüK H, DemetgüL M. Fault diagnosis of bevel gears using neural pattern recognition and MLP neural network algorithms[J]. International Journal of Precision Engineering and Manufacturing, 2020, 21(5): 843-856.
3	Souahlia S, Bacha K, Chaari A. MLP neural network-based decision for power transformers fault diagnosis using an improved combination of rogers and doernenburg ratios DGA[J]. International Journal of Electrical Power & Energy Systems, 2012, 43(1): 1346-1353.
4	Waqar T, Demetgul M. Thermal analysis MLP neural network based fault diagnosis on worm gears[J]. Measurement, 2016, 86: 56-66.
5	Dibaj A, Ettefagh M M, Hassannejad R, et al. A hybrid fine-tuned VMD and CNN scheme for untrained compound fault diagnosis of rotating machinery with unequal-severity faults[J]. Expert Systems with Applications, 2020, 167: 114094.
6	Wang D, Guo Q, Song Y, et al. Application of multiscale learning neural network based on CNN in bearing fault diagnosis[J]. Journal of Signal Processing Systems for Signal, Image, and Video Technology, 2019, 91(10): 1205-1217.
7	Shao S, Yan R, Lu Y, et al. DCNN-based multi-signal induction motor fault diagnosis[J]. IEEE Transactions on Instrumentation and Measurement, 2020, 69(6): 2658-2669.
8	Huang T, Zhang Q, Tang X, et al. A novel fault diagnosis method based on CNN and LSTM and its application in fault diagnosis for complex systems[J]. Artificial Intelligence Review, 2021(5): 1-27.
9	Li X, Li J, Zhao C, et al. Early gear pitting fault diagnosis based on bi-directional LSTM[C]∥Prognostics and System Health Management Conference, Qingdao, 2019:1-5.
10	Han Y, Qi W, Ding N, et al. Short-time wavelet entropy integrating improved LSTM for fault diagnosis of modular multilevel converter[J]. IEEE Transactions on Cybernetics, 2021(99): 1-9.
11	Levent E, Turker I, Serkan K. A generic intelligent bearing fault diagnosis system using compact adaptive 1D CNN classifier[J]. Journal of Signal Processing Systems, 2019, 91: 179-189.
12	Nian-Long G U, Hao P, Peng H E. Bearing fault diagnosis method based on EMD-CNNs[J]. DEStech Transactions on Computer Science and Engineering, 2017, 34: 466-473.
13	Zhang J, Xu B, Wang Z, et al. An FSK-MBCNN based method for compound fault diagnosis in wind turbine gearboxes[J]. Measurement, 2020, 172(6): 108933.
14	杜先君, 贾亮亮. 基于优化堆叠降噪自编码器的滚动轴承故障诊断[J]. 吉林大学学报: 工学版.DOI: 10.13229/j.cnki.jdxbgxb20210415. doi: 10.13229/j.cnki.jdxbgxb20210415
	Du Xian-jun,Jia Liang-liang. Fault diagnosis of rolling bearing based on optimized stacked denoising auto encoders[J]. Journal of Jilin University(Engineering and Technology Edition).DOI: 10.13229/j.cnki.jdxbgxb20210415. doi: 10.13229/j.cnki.jdxbgxb20210415
15	院老虎, 连冬杉, 张亮, 等. 基于密集连接卷积网络和支持向量机的飞行器机械部件故障诊断[J]. 吉林大学学报: 工学版, 2021, 51(5): 1635-1641.
	Yuan Lao-hu,Lian Dong-shan,Zhang Liang, et al. Fault diagnosis of key mechanical components of aircraft based on densenet and support vector machine[J]. Journal of Jilin University(Engineering and Technology Edition), 2021, 51(5): 1635-1641.
16	Russakovsky O, Deng J, Su H, et al. Imagenet large scale visual recognition challenge[J]. International Journal of Computer Vision, 2015, 115: 211-252.
17	Lu L, Wang Z G. Encoding temporal markov dynamics in graph for time series visualization[J]. Association for the Advancement of Artificial Intelligence, 2016, 78: 07273.
18	Jing L, Zhao M, Li P, et al. A convolutional neural network based feature learning and fault diagnosis method for the condition monitoring of gearbox[J]. Measurement, 2017, 111: 1-10.

相关文章 15

[1]	高文志,王彦军,王欣伟,张攀,李勇,董阳. 基于卷积神经网络的柴油机失火故障实时诊断[J]. 吉林大学学报(工学版), 2022, 52(2): 417-424.
[2]	王进花,胡佳伟,曹洁,黄涛. 基于自适应变分模态分解和集成极限学习机的滚动轴承多故障诊断[J]. 吉林大学学报(工学版), 2022, 52(2): 318-328.
[3]	董绍江,朱朋,裴雪武,李洋,胡小林. 基于子领域自适应的变工况下滚动轴承故障诊断[J]. 吉林大学学报(工学版), 2022, 52(2): 288-295.
[4]	罗巍,卢博,陈菲,马腾. 基于PSO-SVM及时序环节的数控刀架故障诊断方法[J]. 吉林大学学报(工学版), 2022, 52(2): 392-399.
[5]	宋林,王立平,吴军,关立文,刘知贵. 基于信息物理融合和数字孪生的可靠性分析[J]. 吉林大学学报(工学版), 2022, 52(2): 439-449.
[6]	邓飞跃,吕浩洋,顾晓辉,郝如江. 基于轻量化神经网络Shuffle⁃SENet的高速动车组轴箱轴承故障诊断方法[J]. 吉林大学学报(工学版), 2022, 52(2): 474-482.
[7]	张龙,徐天鹏,王朝兵,易剑昱,甄灿壮. 基于卷积门控循环网络的齿轮箱故障诊断[J]. 吉林大学学报(工学版), 2022, 52(2): 368-376.
[8]	陈晓雷,孙永峰,李策,林冬梅. 基于卷积神经网络和双向长短期记忆的稳定抗噪声滚动轴承故障诊断[J]. 吉林大学学报(工学版), 2022, 52(2): 296-309.
[9]	刘桂霞,裴志尧,宋佳智. 基于深度学习的蛋白质⁃ATP结合位点预测[J]. 吉林大学学报(工学版), 2022, 52(1): 187-194.
[10]	曲优,李文辉. 基于锚框变换的单阶段旋转目标检测方法[J]. 吉林大学学报(工学版), 2022, 52(1): 162-173.
[11]	张杰,景雯,陈富. 基于被动分簇算法的即时通信网络协议漏洞检测[J]. 吉林大学学报(工学版), 2021, 51(6): 2253-2258.
[12]	欧阳丹彤,张必歌,田乃予,张立明. 结合格局检测与局部搜索的故障数据缩减方法[J]. 吉林大学学报(工学版), 2021, 51(6): 2144-2153.
[13]	董丽丽,杨丹,张翔. 基于深度学习的大规模语义文本重叠区域检索[J]. 吉林大学学报(工学版), 2021, 51(5): 1817-1822.
[14]	院老虎,连冬杉,张亮,刘义. 基于密集连接卷积网络和支持向量机的飞行器机械部件故障诊断[J]. 吉林大学学报(工学版), 2021, 51(5): 1635-1641.
[15]	兰凤崇,李继文,陈吉清. 面向动态场景复合深度学习与并行计算的DG-SLAM算法[J]. 吉林大学学报(工学版), 2021, 51(4): 1437-1446.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed

数据集名称	网络结构	诊断率	最大诊断率
CWRU 1797	Conv1D	0.9966（±0.0016）	0.9985
	MTF-ResNet	0.9923 （±0.0035）	0.9984
	2D-ResNet	0.9312（±0.0306）	0.9766
	M2TF-ResNet	0.9990（±0.0009）	1.0000
CWRU 1772	Conv1D	0.9987（±0.0009）	1.0000
	MTF-ResNet	0.9799（±0.0124）	0.9967
	2D-ResNet	0.9317（±0.0314）	0.9790
	M2TF-ResNet	0.9972（±0.0029）	1.0000
CWRU 1750	Conv1D	0.9985（±0.0007）	0.9996
	MTF-ResNet	0.9633（±0.0264）	0.9848
	2D-ResNet	0.9279（±0.0223）	0.9589
	M2TF-ResNet	0.9974（±0.0024）	1.0000
CWRU 1730	Conv1D	0.9981（±0.0005）	0.9988
	MTF-ResNet	0.9755（±0.0229）	0.9944
	2D-ResNet	0.9094（±0.0270）	0.9511
	M2TF-ResNet	0.9931（±0.0061）	1.0000

一种基于多通道马尔可夫变迁场的故障诊断方法

A fault diagnosis method based on multi Markov transition field

RICH HTML

PDF (PC)

摘要/Abstract

引用本文

使用本文

图/表 7

参考文献 18

相关文章 15

Metrics

本文评价

推荐阅读 0