作者：郝伟¹, 丁昆², 暴长春¹, 贺婷婷¹, 陈仰辉³, 张楷²

作者单位：1. 中车青岛四方机车车辆股份有限公司，山东 青岛 266111;
2. 西南交通大学机械工程学院，四川 成都 610031;
3. 西南交通大学唐山研究院，河北 唐山 063000

关键词：轴承故障诊断;小样本;关系网络;多尺度卷积网络

摘要：

尽管工业条件下可获取大量轴承状态监测数据，但其价值密度低且多为正常状态，可利用的不同类型故障数据较少。针对少样本条件下难以实现高准确率轴承故障诊断的问题，提出一种基于多尺度卷积关系网络的轴承故障诊断方法。该方法首先利用关系网络建立已标记样本之间的对比关系模型；其次，在网络的第一层利用多个大小不同卷积核提取特征并进行特征融合，以增强模型在数据稀缺的条件下对丰富性和互补性故障特征的提取能力；此外，考虑交叉熵损失函数，以提升模型对不同故障类型中判别性特征的提取能力。在帕德博恩大学轴承数据集下，仅利用50条样本训练模型，所提方法相较于WDCNN、SECNN、孪生网络、原型网络和关系网络对1000条无标记样本的平均测试准确率分别提升33.66%，28.63%，7.62%，7.82%和4.21%。此外，对机车轴承数据集添加SNR为-1 dB的高斯白噪声以模拟强噪声干扰环境，所提方法仅利用20条训练样本对1200条测试样本达到89.83%的较高诊断精度。实验结果显示，在小样本训练条件下，所提方法能够有效提升模型的泛化、抗噪和辨识能力。

Bearing fault diagnosis method based on multi-scale convolutional relation network with limited samples
HAO Wei¹, DING Kun², BAO Changchun¹, HE Tingting¹, CHEN Yanghui³, ZHANG Kai²
1. CRRC Qingdao Sifang Co., Ltd., Qingdao 266111, China;
2. School of Mechanical Engineering, Southwest Jiaotong University, Chengdu 610031, China;
3. Tangshan Research Institute, Southwest Jiaotong University, Tangshan 063000, China
Abstract: Although a large amount of bearing condition monitoring data can be obtained under industrial conditions, its value density is low and mostly normal. There are fewer different types of fault data available. A rolling bearing fault diagnosis method based on a multi-scale convolutional relation network is proposed to address the difficult problem of achieving high accuracy with limited samples. Firstly, a frame based on a relation network is designed to establish a comparative relationship between the labeled samples. Secondly, the first layer of the designed frame is improved by multiple convolutional kernels with different sizes to enhance the extraction and fusion of rich and complementary fault features under data scarcity. Thirdly, the cross-entropy loss function is considered to enhance the model’s extraction capability for discriminative features in different fault types. With the Paderborn University bearing data, the proposed method improves the average test accuracy by 33.66%, 28.63%, 7.62%, 7.82% and 4.21% compared to WDCNN, SECNN, Siamese network, Prototypical network, and Relation network for 1000 unlabeled samples using only 50 samples to train the model, respectively. Additionally, the proposed method achieves a high diagnostic accuracy of 89.83% for 1200 test samples using only 20 training samples by adding white Gaussian noise with SNR of -1 dB to the locomotive bearing dataset to simulate a strong noise interference environment. The experimental illustrates show that the proposed method can effectively improve the model’s ability of generalization, anti-noise and discrimination under limited training samples.
Keywords: bearing fault diagnosis;limited samples;relation network;multi-scale convolutional network
2024, 50(3):160-168  收稿日期: 2022-07-06;收到修改稿日期: 2022-11-07
基金项目: 中央高校基本科研业务费（2682022CX006）；国家重点研发计划项目（2021YFB3400700）
作者简介: 郝伟(1985-),女,山东聊城市人,高级工程师,博士,研究方向为轨道交通信息技术。
参考文献
[1] 赵磊, 张永祥, 朱丹宸. 复杂装备滚动轴承的故障诊断与预测方法研究综述[J]. 中国测试, 2020, 46(3): 17-25.
ZHAO L, ZHANG Y X, ZHU D C. A review of research on fault diagnosis and prediction methods for complex equipment rolling bearings [J].China Measurement ＆ Test,2020, 46 (3): 17-25.
[2] ZHANG W, PENG G, LI C, et al. A new deep learning model for fault diagnosis with good anti-noise and domain adaptation ability on raw vibration signals[J]. Sensors, 2017, 17(2): 425.
[3] 邓佳林, 邹益胜, 张笑璐, 等. 一种改进CNN在轴承故障诊断中的应用[J]. 现代制造工程, 2020(4): 142-147.
DENG J L, ZOU Y S, ZHANG X L, et al, An improved application of CNN in bearing fault diagnosis [J].Modern Manufacturing Engineering, 2020 (4): 142-147.
[4] 周林春, 陈春俊. 复数卷积神经网络滚动轴承故障诊断研究[J]. 中国测试, 2020, 46(11): 109-115.
ZHOU L C, CHEN C J. Research on fault diagnosis of rolling bearings using complex convolutional neural networks [J].China Measurement ＆ Test, 2020, 46 (11): 109-115.
[5] 杨健, 李立新, 廖晨茜, 等. 面向滚动轴承故障诊断的改进对抗迁移学习算法研究[J]. 中国测试, 2021, 47(9): 15-19+40.
YANG J, LI L X, LIAO C Q, et al.Research on improved adversarial transfer learning algorithm for fault diagnosis of rolling bearings [J].China Measurement＆ Test, 2021, 47 (9): 15-19+40.
[6] KOCH G, ZEMEL R, SALAKHUTDINOV R. Siamese neural networks for one-shot image recognition[C]//ICML deep learning workshop. 2015, 2: 1-30.
[7] 刘岱, 常东润, 孙习习, 等. 基于卷积孪生神经网络的滚动轴承故障定位方法[J]. 机电工程, 2022, 39(3): 309-316.
LIU D, CHANG D R, SUN X X, et al. A rolling bearing fault localization method based on convolutional twin neural network [J].Journal of Mechanical ＆ Electrical Engineering, 2022, 39 (3): 309-316.
[8] SNELL J, SWERSKY K, ZEMEL R. Prototypical networks for few-shot learning[C]//Advances in neural information processing systems, 2017, 2017-Decem: 4078-4088.
[9] 余浩帅, 汤宝平, 张楷, 等. 小样本下混合自注意力原型网络的风电齿轮箱故障诊断方法[J]. 中国机械工程, 2021, 32(20): 2475-2481.
YU H S, TANG B P, ZHANG K, et al. A fault diagnosis method for wind power gearbox using a mixed self attention prototype network under small sample size [J].China Mechanical Engineering, 2021, 32 (20): 2475-2481.
[10] SUNG F, YANG Y, ZHANG L, et al. Learning to compare: relation network for few-shot learning[C]//Proceedings of the IEEE conference on computer vision and pattern recognition. 2018: 1199-1208.
[11] RUAN H, WANG Y, LI X, et al. A relation-based semisupervised method for gearbox fault diagnosis with limited labeled samples[J]. IEEE Transactions on Instrumentation and Measurement, 2021, 70: 3510013.
[12] LU N, HU H, YIN T, et al. Transfer relation network for fault diagnosis of rotating machinery with small data[J]. IEEE Transactions on Cybernetics,52(11): 11927-11941.
[13] 许子非, 金江涛, 李春. 基于多尺度卷积神经网络的滚动轴承故障诊断方法[J]. 振动与冲击, 2021, 40(18): 212-220.
XU Z F, JIN J T, LI C. A fault diagnosis method for rolling bearings based on multi-scale convolutional neural networks [J].Journal of Vibration and Shock, 2021, 40 (18): 212-220.
[14] QIAO H, WANG T, WANG P, et al. An adaptive weighted multi-scale convolutional neural network for rotating machinery fault diagnosis under variable operating conditions[J]. IEEE Access, 2019, 7: 118954-118964.
[15] ZOU Y, SHI K, LIU Y, et al. Rolling bearing transfer fault diagnosis method based on adversarial variational autoencoder network[J]. Measurement Science and Technology, 2021, 32(11): 115017.
[16] KIMOTHO J K, LESSMEIER C, SEXTRO W, et al. Condition monitoring of bearing damage in electromechanical drive systems by using motor current signals of electric motors: a benchmark data set for data-driven classification[C]//PHM Society European Conference. 2016, 3(1): 152-156.