Research on Fault Diagnosis Method of Deep Adversarial Transfer Learning
-
摘要: 针对实验室环境下容易获取大量有标签故障类型数据，而在实际工况条件下很难或无法获取大量带标签数据的问题，提出机械设备故障的深度对抗迁移诊断方法（MAAN）。该方法将实验室环境中积累的故障诊断知识迁移应用于工程实际装备，通过融合时域与频域数据获取更全面的故障信息，在特征提取层利用残差网络深度提取故障特征，对抗层采用最大化域分类损失用于对齐源域与目标域的边缘分布和条件概率分布，最小化类别预测损失用于机械设备的故障分类实现无监督迁移学习。实验结果表明，此模型在无标签的目标数据集中有较高的分类精度，在一定条件下可以有效解决数据集缺少标签的难题，即实现机械故障诊断的智能诊断。Abstract: Aiming at the problem that it is easy to obtain a large amount of labeled fault type data in a laboratory environment, but it is difficult or impossible to obtain a large amount of labeled data under actual working conditions, a deep anti-migration diagnosis method (MAAN) for mechanical equipment failure is proposed. This method transfers the fault diagnosis knowledge accumulated in the laboratory environment to the actual engineering equipment, through the fusion of time domain and frequency domain data to obtain more comprehensive fault information; in the feature extraction layer, the residual network is used to deeply extract fault features. The adversarial layer uses maximizing domain classification loss to align the marginal distribution and conditional probability distribution between the source domain and target domain, and minimizing the category prediction loss to implement unsupervised migration learning for fault classification of mechanical equipment. Experimental results show that this method and its network models have high classification accuracy in unlabeled target data sets. Under certain conditions, this method can effectively solve the problem of lack of labels in data sets, that is, realize intelligent diagnosis of mechanical fault diagnosis.
-
Key words:
- migration diagnosis /
- deep learning /
- adversarial network /
- unsupervised learning
-
表 3 轴承数据集
滚动轴承数据 负载 健康状态 样本个数 内圈故障 1000 XJTU-SU(A) 11 kN 外圈故障 1000 正常 1000 内圈故障 1000 CWRU(B) 0 外圈故障 1000 （0.18） 正常 1000 
下载: 导出CSV
表 4 CWRU轴承故障数据集
负载/HP 损伤直径/mm 健康状态 样本个数
0/ 1/
2/ 30 正常 1000
0.07内圈故障 1000 外圈故障 1000 滚动体故障 1000
0.14内圈故障 1000 外圈故障 1000 滚动体故障 1000
0.21内圈故障 1000 外圈故障 1000 滚动体故障 1000
下载: 导出CSV
-
[1] 雷亚国, 贾峰, 孔德同, 等. 大数据下机械智能故障诊断的机遇与挑战[J]. 机械工程学报, 2018, 54(5): 94-104 doi: 10.3901/JME.2018.05.094LEI Y G, JIA F, KONG D T, et al. Opportunities and challenges of machinery intelligent fault diagnosis in big data era[J]. Journal of Mechanical Engineering, 2018, 54(5): 94-104 (in Chinese) doi: 10.3901/JME.2018.05.094 [2] YOSINSKI J, CLUNE J, BENGIO Y, et al. How transferable are features in deep neural networks?[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal, Canada: MIT Press, 2014: 3320-3328 [3] 康守强, 胡明武, 王玉静, 等. 基于特征迁移学习的变工况下滚动轴承故障诊断方法[J]. 中国电机工程学报, 2019, 39(3): 764-772KANG S Q, HU M W, WANG Y J, et al. Fault diagnosis method of a rolling bearing under variable working conditions based on feature transfer learning[J]. Proceedings of the CSEE, 2019, 39(3): 764-772 (in Chinese) [4] TONG Z, LI W, ZHANG B, et al. Bearing fault diagnosis under variable working conditions based on domain adaptation using feature transfer learning[J]. IEEE Access, 2018, 6: 76187-76197 doi: 10.1109/ACCESS.2018.2883078 [5] 雷亚国, 杨彬, 杜兆钧, 等. 大数据下机械装备故障的深度迁移诊断方法[J]. 机械工程学报, 2019, 55(7): 1-8 doi: 10.3901/JME.2019.07.001LEI Y G, YANG B, DU Z J, et al. Deep transfer diagnosis method for machinery in big data era[J]. Journal of Mechanical Engineering, 2019, 55(7): 1-8 (in Chinese) doi: 10.3901/JME.2019.07.001 [6] 张根保, 李浩, 冉琰, 等. 一种用于轴承故障诊断的迁移学习模型[J]. 吉林大学学报, 2020, 50(5): 1617-1626ZHANG G B, LI H, RAN Y, et al. A transfer learning model for bearing fault diagnosis[J]. Journal of Jilin University, 2020, 50(5): 1617-1626 (in Chinese) [7] 龙明盛. 迁移学习问题与方法研究[D]. 北京: 清华大学, 2014LONG M S. Transfer learning: problems and methods[D]. Beijing: Tsinghua University, 2014 (in Chinese) [8] ARJOVSKY M, CHINTALA S, BOTTOU L. Wasserstein GAN[Z]. arXiv preprint arXiv: 1701.07875, 2017 [9] GOODFELLOW I J, POUGET-ABADIE J, MIRZA M, et al. Generative adversarial nets[C]//Proceedings of the 27th International Conference on Neural Information Processing Systems. Montreal, Canada: MIT Press, 2014: 2672-2680 [10] PAN S J, YANG Q. A survey on transfer learning[J]. IEEE Transactions on Knowledge and Data Engineering, 2010, 22(10): 1345-1359 doi: 10.1109/TKDE.2009.191 [11] HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Las Vegas: IEEE, 2016: 770-778 [12] GANIN Y, LEMPITSKY V. Unsupervised domain adaptation by backpropagation[C]//Proceedings of the 32nd International Conference on International Conference on Machine Learning. Lille, France: JMLR. org, 2015: 1180-1189 [13] YU C H, WANG J D, CHEN Y Q, et al. Transfer learning with dynamic adversarial adaptation network[C]//2019 IEEE International Conference on Data Mining (ICDM). Beijing, China: IEEE, 2019: 778-786 [14] 雷亚国, 韩天宇, 王彪, 等. XJTU-SY滚动轴承加速寿命试验数据集解读[J]. 机械工程学报, 2019, 55(16): 1-6 doi: 10.3901/JME.2019.16.001LEI Y G, HAN T Y, WANG B, et al. XJTU-SY rolling element bearing accelerated life test datasets: a tutorial[J]. Journal of Mechanical Engineering, 2019, 55(16): 1-6 (in Chinese) doi: 10.3901/JME.2019.16.001 [15] TZENG E, HOFFMAN J, ZHANG N, et al. Deep domain confusion: maximizing for domain invariance[J]. arXiv preprint arXiv: 1412.3474, 2014 [16] GANIN Y, USTINOVA E, AJAKAN H, et al. Domain-adversarial training of neural networks[J]. The Journal of Machine Learning Research, 2016, 17(1): 2096-2030 -
点击查看大图
图(8) / 表(4)
计量
- 文章访问数: 366
- HTML全文浏览量: 68
- PDF下载量: 25
- 被引次数: 0
