针对滚动轴承故障样本稀缺、振动特征提取困难导致故障诊断准确率低的难题,提出一种基于时频增强的滚动轴承少样本故障诊断方法。首先,对滚动轴承一维振动信号进行重叠采样,利用连续小波变换对采样信号段进行时频域特征映射,构造二维时频矩阵;其次,通过深度卷积生成对抗网络对真实时频样本进行训练后,将生成时频样本加入到训练集中;然后,采用时序卷积网络融合深层次的时频域特征;最后,构建Softmax分类器输出与故障类别对应的状态。仿真实验结果表明,在仅有10个训练样本的条件下,该方法在凯斯西储大学滚动轴承数据集中的诊断准确率均值达91.00%,相较未经时频增强的方法提高了7.56%,并利用实测数据验证了时频增强方法能够显著提升少样本情形下的故障诊断准确率。
Abstract
Aiming at the problem of low fault diagnosis accuracy due to the scarcity of fault samples of rolling bearings and the difficulty in extracting vibration features, a fault diagnosis method of rolling bearings with few samples based on time-frequency enhancement is proposed. Firstly, one-dimensional vibration signals of rolling bearings are overlapped and sampled, and a two-dimensional time-frequency matrix is constructed by using continuous wavelet transform to map the sampled signals in time-frequency domain. Secondly, the real time-frequency samples are trained by deep convolutional generative adversarial network, and the generated time-frequency samples are added to the training set. Then, the time-series convolutional network is used to fuse the deep time-frequency domain features. Finally, Softmax classifier is constructed to output states corresponding to fault categories. Simulation experiment results show that under the condition of only 10 training samples, the method in the diagnosis of rolling bearing at case western reserve university data set average accuracy of 91.00%, compared with an enhanced time-frequency method is increased by 7.56%, and the time-frequency method is validated by the measured data can significantly improve the fault diagnosis accuracy of less sample circumstances.
关键词
计量学 /
滚动轴承 /
故障诊断 /
时频增强 /
少样本
Key words
metrology /
rolling bearing /
fault diagnosis /
time-frequency enhancement /
less sample
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
［1］ViolaJ, Chen Y Q, Wang J.FaultFace: Deep convolutional generative adversarial network (DCGAN) based ball-bearing failure detection method［J］. Information Sciences, 2021, 542(1): 195-211.
［2］李继猛, 王慧, 李铭, 等. 基于改进的自适应无参经验小波变换的滚动轴承故障诊断［J］. 计量学报, 2020, 41(6): 710-716.
Li J M, Wang H, Li M, et al. Rolling Bearing Fault Diagnosis Based on Improved Adaptive Parameterless Empirical Wavelet Transform［J］. Acta Metrologica Sinica, 2020, 41(6): 710-716.
［3］Wang B J, Zhang X L, Xing S, et al. Sparse representation theory for support vector machine kernel function selection and its application inhigh-speed bearing fault diagnosis ［J］. ISA Transactions, 2021, 118: 207-218.
［4］Xia T B, Zhuo P C, Xiao L, et al. Multi-stage fault diagnosis framework for rolling bearing based on OHF Elman AdaBoost-Bagging algorithm［J］. Neurocomputing, 2021, 433(4): 237-251.
［5］Li X, Yang Y, Pan H Y, et al. Non-parallel least squares support matrix machine for rolling bearing fault diagnosis［J］. Mechanism and Machine Theory, 2020, 145(3): 103676.
［6］文成林, 吕菲亚. 基于深度学习的故障诊断方法综述［J］. 电子与信息学报, 2020, 42(1): 234-248.
Wen C L, Lv F Y. Review of Fault Diagnosis Methods Based on Deep Learning ［J］. Journal of Electronics & Information Technology, 2020, 42(1): 234-248.
［7］Wang X, Mao D X, Li X D. Bearing fault diagnosis based on vibro-acoustic data fusion and 1D-CNN network［J］. Measurement, 2020, 173(3): 108518.
［8］Cheng Y, Lin M, Wu J, et al. Intelligent fault diagnosis of rotating machinery based on continuous wavelet transform-local binary convolutional neural network［J］. Knowledge-Based Systems, 2021, 216(3): 106796.
［9］Dong Y J, Li Y Q, Zheng H L, et al. A new dynamic model and transfer learning based intelligent fault diagnosis framework for rolling element bearings race faults: Solving the small sample problem ［J］. ISA Transactions, 2022, 121: 327-348.
［10］何强, 唐向红, 李传江, 等. 负载不平衡下小样本数据的轴承故障诊断［J］. 中国机械工程, 2021, 32(10): 1164-1171+1180.
He Q, Tang X H, Li C J, et al. Bearing Fault Diagnosis with Small Sample Data under Load Unbalance ［J］. China Mechanical Engineering, 2021, 32(10): 1164-1171+1180.
［11］Han T, Liu C, Yang W G, et al. A novel adversarial learning framework in deep convolutional neural network for intelligent diagnosis of mechanical faults［J］. Knowledge-Based Systems, 2019, 165(2): 474-487.
［12］陈维兴, 崔朝臣, 李小菁, 等. 基于多种小波变换的一维卷积循环神经网络的风电机组轴承故障诊断［J］. 计量学报, 2021, 42(5): 615-622.
Chen W X, Cui C C, Li X J, et al. Bearing Fault Diagnosis of Wind Turbine Based on Multi-wavelet-1D Convolutional LSTM［J］. Acta Metrologica Sinica, 2021, 42(5): 615-622.
［13］Goodfellow I, Pouget-abadie J, Mirza M, et al. Generative adversarial nets［C］//Advances in Neural Information Processing Systems, 2014: 2672-2680.
［14］董绍江, 裴雪武, 吴文亮, 等. 基于多层降噪技术及改进卷积神经网络的滚动轴承故障诊断方法［J］. 机械工程学报, 2021, 57(1): 148-156.
Dong S J, Pei X W, Wu W L, et al. Rolling Bearing Fault Diagnosis Method Based on Multi-layer Denoising Technology and Improved Convolutional Neural Network ［J］. Journal of Mechanical Engineering, 2021, 57(1): 148-156.
［15］陈剑, 黄凯旋, 吕伍佯, 等. 基于VMD和卷积神经网络的变工况轴承故障诊断方法［J］. 计量学报, 2021, 42(7): 892-897.
Chen J, Huang K X, Lü W Y, et al. Bearing Fault Diagnosis Method Based on VMD and Convolutional Neural Network Undervarying Operation Conditions［J］. Acta Metrologica Sinica, 2021, 42(7): 892-897.
［16］Zhao B X, Yuan Q. Improved generative adversarial network for vibration-based fault diagnosis with imbalanced data［J］. Measurement, 2021, 169(2): 108522.
基金
重庆市自然科学基金(cstc2021jcyj-msxmX0330)
PDF(5641 KB)
