为解决实际应用中风力机滚动轴承故障训练样本严重不足的问题,提出一种基于改进残差神经网络与迁移学习的小样本滚动轴承故障诊断模型。首先,该模型将挤压与激励网络嵌入到一维残差神经网络中,增加了模型的特征提取能力;其次,使用源域数据对所搭建改进残差神经网络模型进行训练,确定结构和参数,并使用L2正则化和Dropout机制抑制过拟合;然后,引入迁移学习,冻结使用源域数据训练好的部分模型参数,使用少量目标域数据对模型的全连接层参数进行微调;最后,对不同故障的样本进行分类。该方法在凯斯西储大学轴承数据集和本实验室轴承数据集上进行实验验证,实验结果表明：在不同实验条件下,所提方法与其他方法的计算结果进行比较,其均有更高的故障诊断准确度和更强的泛化能力。
Abstract
In order to solve the problem of few fault training samples of wind turbines rolling bearings in practical applications, a novel rolling bearing fault diagnosis model based on improved residual neural network and transfer learning under small samples condition is proposed. Firstly, the squeeze-and-excitation networks is embedded into the one-dimensional residual neural network, which increases the feature extraction ability of the model. Secondly, the improved residual neural network model is trained with source domain data to determine the structure and parameters, and L2 regularization method and Dropout mechanism are used to suppress overfitting. Then, transfer learning is introduced to freeze some model parameters trained with source domain data, and fine-tune the full connection layer parameters of the model by using a small amount of target domain data. Finally, the samples of different faults are classified. This method is verified by experiments on the bearing data set of Case Western Reserve University and the bearing data set of our laboratory. The experimental results show that under different experimental conditions, the proposed method has higher fault diagnosis accuracy and stronger generalization ability than that of compared methods.
关键词
风力机 /
滚动轴承 /
故障诊断 /
迁移学习 /
挤压与激励网络 /
小样本
Key words
wind turbines /
rolling bearings /
fault diagnosis /
transfer learning /
squeeze-and-excitation networks /
small sample
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] CHEN Z Q, DENG S C, CHEN X D, et al.Deep neural networks-based rolling bearing fault diagnosis[J]. Microelectronics reliability, 2017, 75: 327-333.
[2] 向玲, 李营. 风力机组滚动轴承复合故障诊断研究[J]. 太阳能学报, 2021, 42(3): 90-97.
XIANG L,LI Y.Research on composite fault diagnosis of wind turbine rolling bearings[J]. Acta energiae solaris sinica, 2021, 42(3): 90-97.
[3] 张文龙, 胡天亮, 王艳洁, 等. 云/边缘协同的轴承故障诊断方法[J]. 计算机集成制造系统, 2020, 26(3): 589-599.
ZHANG W L, HU T L, WANG Y J, et al.A cloud/edge collaborated bearing fault diagnosis method[J]. Computer integrated manufacturing systems, 2020, 26(3): 589-599.
[4] 赵春华, 胡恒星, 陈保家, 等. 基于深度学习特征提取和WOA-SVM状态识别的轴承故障诊断[J]. 振动与冲击, 2019, 38(10): 31-37, 48.
ZHAO C H, HU H X, CHEN B J, et al.Bearing fault diagnosis based on the deep learning feature extraction and WOA-SVM state recognition[J]. Journal of vibration and shock, 2019, 38(10): 31-37, 48.
[5] LIU H, ZHOU J Z, ZHENG Y, et al.Fault diagnosis of rolling bearings with recurrent neural network based autoencoders[J]. ISA transactions, 2018, 77: 167-178.
[6] 余萍, 曹洁. 优化堆叠降噪自动编码器滚动轴承故障诊断[J]. 太阳能学报, 2021, 42(11): 307-314.
YU P, CAO J.Optimized stacked denoising auto-encoders (SDAE)-based fault diagnosis of rolling bearing[J]. Acta energiae solaris sinica, 2021, 42(11): 307-314.
[7] PENG B S, XIA H, LYU X Z, et al.An intelligent fault diagnosis method for rotating machinery based on data fusion and deep residual neural network[J]. Applied intelligence, 2022, 52: 3051-3065.
[8] 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(工学版), 2020, 50(5): 1617-1626.
ZHANG G B, LI H, RAN Y, et al.A transfer learning model for bearing fault diagnosis[J]. Journal of Jilin University(engineering and technology edition), 2020, 50(5): 1617-1626.
[9] XUE T, WU D H, WANG H G, et al.Research on application of transfer learning in equipment fault diagnosis[J]. Journal of physics: conference series, 2021, 1986(1): 012099.
[10] WEISS K, KHOSHGOFTAAR T M, WANG D D.A survey of transfer learning[J]. Journal of big data, 2016, 3(1): 9-48.
[11] 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): 764-772.
KANG 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.
[12] WU J Y, ZHAO Z B, SUN C, et al.Few-shot transfer learning for intelligent fault diagnosis of machine[J]. Measurement, 2020, 166: 108202.
[13] CHEN Z H, CEN J, XIONG J B.Rolling bearing fault diagnosis using time-frequency analysis and deep transfer convolutional neural network[J]. IEEE access, 2020, 8: 150248-150261.
[14] ZHENG Z, FU J M, LU C Q, et al.Research on rolling bearing fault diagnosis of small dataset based on a new optimal transfer learning network[J]. Measurement, 2021, 177: 109285.
[15] HE K M, ZHANG X Y, REN S Q.Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas, USA, 2016: 770-778.
[16] HU J, SHEN L, SUN G, et al.Squeeze-and-excitation networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition(CVPR), Salt Lake City, USA, 2018: 7132-7141.
[17] PAN S J, YANG Q.A survey on transfer learning[J]. IEEE transactions on knowledge and data engineering, 2010, 22(10): 1345-1359.
[18] LAPARO K A. Bearing vibration data set[EB/OL]. http:// www.eecs.cwru.edu/laboratory/bearing/download.htm.
基金
国家自然科学基金（51465035）; 甘肃省自然科学基金（20JR5RA466）
PDF(1767 KB)
