一种用于轴承故障诊断的二维卷积神经网络优化方法
|
摘要
轴承智能故障诊断是机械大数据状态监测的热门研究领域,传统的数据驱动故障诊断方法对基于信号处理的特征提取环节极为耗时,并且对专家经验要求高,为消除其带来的参数预定义影响,实现快速特征提取的同时提高识别率,在研究一维卷积神经网络故障诊断方法的基础上,提出一种用于轴承故障诊断的二维卷积神经网络优化方法。该方法引入了一种新的数据预处理方式,将原始时域信号数据转换成二维灰度图像来提取转换后的图像特征,消除手工特征的影响;同时,在验证分类前对实验采集故障数据集添加了降噪处理,并对卷积神经网络梯度下降算法进行参数自适应学习率优化。仿真与实验结果表明,所提出的二维优化卷积神经网络故障诊断方法在选取64×64的信号–图片转换格式下,AMSGrad算法能将故障预测模型的准确度提升至98%,训练速度更快,同时具有更高的分类准确性和抗噪性能,使其在实际转速范围内能达到损失小于5%的识别准确率。
Intelligent bearing fault diagnosis is a hot research field of mechanical big data condition monitoring. The traditional data-driven fault diagnosis method is extremely time-consuming and requires high expert experience for signal extraction-based feature extraction. In order to eliminate the pre-defined effects of parameters and improve the recognition rate while improving the feature extraction, this paper proposed a two-dimensional convolutional neural network optimization method for bearing fault diagnosis based on the research of one-dimensional convolution neural network fault diagnosis method. This method introduced a new data preprocessing method, which converted the original time domain signal data into a two-dimensional gray image to extract the transformed image features, eliminated the influence of manual features,and collects the experimental errors before verifying the classification. The data set adds noise reduction processing and optimizes the parameter adaptive learning rate for the convolutional neural network gradient descent algorithm. The simulation and experimental results show that the proposed two-dimensional optimized convolutional neural network fault diagnosis method is based on the signal-picture conversion format under 64×64. The AMSGrad algorithm can improve the accuracy of the fault prediction model to 98%, and the training speed is faster and higher. Classification accuracy and anti-noise performance can achieve a recognition accuracy of less than 5% loss in the actual speed range.
Intelligent bearing fault diagnosis is a hot research field of mechanical big data condition monitoring. The traditional data-driven fault diagnosis method is extremely time-consuming and requires high expert experience for signal extraction-based feature extraction. In order to eliminate the pre-defined effects of parameters and improve the recognition rate while improving the feature extraction, this paper proposed a two-dimensional convolutional neural network optimization method for bearing fault diagnosis based on the research of one-dimensional convolution neural network fault diagnosis method. This method introduced a new data preprocessing method, which converted the original time domain signal data into a two-dimensional gray image to extract the transformed image features, eliminated the influence of manual features,and collects the experimental errors before verifying the classification. The data set adds noise reduction processing and optimizes the parameter adaptive learning rate for the convolutional neural network gradient descent algorithm. The simulation and experimental results show that the proposed two-dimensional optimized convolutional neural network fault diagnosis method is based on the signal-picture conversion format under 64×64. The AMSGrad algorithm can improve the accuracy of the fault prediction model to 98%, and the training speed is faster and higher. Classification accuracy and anti-noise performance can achieve a recognition accuracy of less than 5% loss in the actual speed range.
引文
[1]刘长良,武英杰,甄成刚.基于变分模态分解和模糊C均值聚类的滚动轴承故障诊断[J].中国电机工程学报,2015,35(13):3358-3365.Liu Changliang,Ying Wujie,Zhen Chenggang.Rolling bearing fault diagnosis based on variational mode decomposition and fuzzy C means clustering[J].Proceedings of the CSEE,2015,35(13):3358-3365(in Chinese).
[2]Gonzalez-Prieto I,Duran M J,Rios-Garcia N,et al.Open-switch fault detection in five-phase induction motor drives using model predictive control[J].IEEETransactions on Industrial Electronics,2018,65(4):3045-3055.
[3]杨明,柴娜,李广.等.基于电机驱动系统的齿轮故障诊断方法对比研究[J].电工技术学报,2016,31(19):132-140.Yang Ming,Chai Na,Li Guang.et al.A comparative study of gear fault diagnosis methods based on the motor drive system[J].Transactions of China Electrotechnical Society,2016,31(19):132-140(in Chinese).
[4]王臻,李承,张舜钦,等.基于改进矩阵束滤波与检测的异步电动机故障诊断新方法[J].中国电机工程学报,2015,35(12):3969-3977.Wang Zhen,Li Cheng,Zhang Shunqin,et al.A new filter and detection approach of fault in asynchronous motor based on improved matrix pencil[J].Proceedings of the CSEE,2015,35(12):3969-3977(in Chinese).
[5]魏书荣,李正茂,符杨,等.计及电流估计差的海上双馈电机定子绕组匝间短路故障诊断[J].中国电机工程学报,2018,38(13):213-219.Wei Shurong,Li Zhengmao,Fu Yang,et al.Inter-turn faults diagnosis on stator windings of offshore wind DFIGs considering current estimated difference[J].Proceedings of the CSEE,2018,38(13):213-219(in Chinese).
[6]袁莉芬,孙业胜,何怡刚,等.基于小波包优选的模拟电路故障特征提取方法[J].电工技术学报,2018,33(11):158-165.Yuan Lifen,Sun Yesheng,He Yigang,et al.Fault feature extraction method for analog circuit based on preferred wavelet packet[J].Transactions of China Electrotechnical Society,2018,33(11):158-165(in Chinese).
[7]Gao Z W,Cecati C,Ding Seven X.A survey of fault diagnosis and fault-tolerant techniques-Part I:Fault diagnosis with model-based and signal-based approaches[J].IEEE Transactions on Industrial Electronics,2015,62(6):3757-3767.
[8]Wang D,Peter W T.Prognostics of slurry pumps based on a moving-wear degradation index and a general sequential Monte Carlo method[J].Mechanical Systems and Signal Processing,2015,56-57:213-229.
[9]周国亮,朱永利,王桂兰.实时大数据处理技术在状态监测领域中的应用[J].电工技术学报,2014,29(S1):432-437.Zhou Guoliang,Zhu Yongli,Wang Guilan.Real-time big data processing technology application in the field of state monitoring[J].Transactions of China Electrotechnical Society,2014,29(S1):432-437(in Chinese).
[10]朱永利,李莉,宋亚奇,等.ODPS平台下的电力设备监测大数据存储与并行处理方法[J].电工技术学报,2017,32(9):199-210.Zhu Yongli,Li Li,Song Yaqi,et al.Storage and parallel processing of big data of power equipment condition monitoring on ODPS platform[J].Transactions of China Electrotechnical Society,2017,32(9):199-210(in Chinese).
[11]AYDMJ T,Duin R P W.Pump failure determination using support vector data description[C]//Proceedings of the19th ACM Conference on Information and Knowledge Management,Toronto,2010.
[12]梅飞,梅军,郑建勇,等.基于KPCA-SVM的断路器故障稳健诊断方法[J].电工技术学报,2014,29(S1):50-58.Mei Fei,Mei Jun,Zheng Jianyong,et al.The robust fault diagnostic method for circuit breaker based on KPCA and SVM[J].Transactions of China Electrotechnical Society,2014,29(S1):50-58(in Chinese).
[13]廖剑,史俊贤,周绍磊,等.基于局部图嵌入加权罚SVM的模拟电路故障诊断方法[J].电工技术学报,2016,31(4):28-35.Liao Jian,Shi Junxian,Zhou Shaolei,et al.Analog circuit fault diagnosis based on local graph embedding weightedpenalty SVM[J].Transactions of China Electrotechnical Society,2016,31(4):28-35(in Chinese).
[14]孙曙光,张强,杜太行,等.基于振动信号的低压万能式断路器分合闸故障程度评估方法的研究[J].中国电机工程学报,2017,37(18):5473-5482.Sun Shuguang,Zhang Qiang,Du Taihang,et al.Study of evaluation method for low voltage conventional circuit breaker switching fault degree based on vibration signal[J].Proceedings of the CSEE,2017,37(18):5473-5482(in Chinese).
[15]史丽萍,汤家升,王攀攀,等.采用最优小波树和改进BP神经网络的感应电动机定子故障诊断[J].电工技术学报,2015,30(24):38-45.Shi Liping,Tang Jiasheng,Wang Panpan,et al.Stator fault diagnosis of induction motors using the optimal wavelet tree and improved BP neural network[J].Transactions of China Electrotechnical Society,2015,30(24):38-45(in Chinese).
[16]杨凌霄,朱亚丽.基于概率神经网络的高压断路器故障诊断[J].电力系统保护与控制,2015,43(10):62-67.Yang Lingxiao,Zhu Yali.High voltage circuit breaker fault diagnosis of probabilistic neural network[J].Power System Protection and Control,2015,43(10):62-67(in Chinese).
[17]Shao H,Jiang H,Zhang X,et al.Rolling bearing fault diagnosis using an optimization deep belief network[J].Measurement Science and Technology,2015,26(11):11502.
[18]Wen L,Gao L,Li X Y.A new deep transfer learning based on sparse auto-encoder for fault diagnosis[J].IEEETransactions on Systems Man and Cybernetics:Systems,2017(99):1-9.
[19]候文擎,叶鸣,李巍华.基于改进堆叠降噪自编码的滚动轴承故障分类[J].机械工程学报,2018,54(7):87-96.Hou Wenqing,Ye Ming,Li Weihua.Rolling element bearing fault classification using improved stacked de-noising auto-encoders[J].Journal of Mechanical Engineering,2018,54(7):87-96(in Chinese).
[20]Ince T,Kiranyaz S,Eren L,et al.Real-time motor fault detection by 1-d convolutional neural networks[J].IEEETransactions on Industrial Electronics,2016,63(11):7067-7075.
[21]Reddi S J,Kale S,Kumar S.On the convergence of adam and beyond[C]//International Conference on Learning Representations(ICLR),Vancouver,Canada,2018.
[2]Gonzalez-Prieto I,Duran M J,Rios-Garcia N,et al.Open-switch fault detection in five-phase induction motor drives using model predictive control[J].IEEETransactions on Industrial Electronics,2018,65(4):3045-3055.
[3]杨明,柴娜,李广.等.基于电机驱动系统的齿轮故障诊断方法对比研究[J].电工技术学报,2016,31(19):132-140.Yang Ming,Chai Na,Li Guang.et al.A comparative study of gear fault diagnosis methods based on the motor drive system[J].Transactions of China Electrotechnical Society,2016,31(19):132-140(in Chinese).
[4]王臻,李承,张舜钦,等.基于改进矩阵束滤波与检测的异步电动机故障诊断新方法[J].中国电机工程学报,2015,35(12):3969-3977.Wang Zhen,Li Cheng,Zhang Shunqin,et al.A new filter and detection approach of fault in asynchronous motor based on improved matrix pencil[J].Proceedings of the CSEE,2015,35(12):3969-3977(in Chinese).
[5]魏书荣,李正茂,符杨,等.计及电流估计差的海上双馈电机定子绕组匝间短路故障诊断[J].中国电机工程学报,2018,38(13):213-219.Wei Shurong,Li Zhengmao,Fu Yang,et al.Inter-turn faults diagnosis on stator windings of offshore wind DFIGs considering current estimated difference[J].Proceedings of the CSEE,2018,38(13):213-219(in Chinese).
[6]袁莉芬,孙业胜,何怡刚,等.基于小波包优选的模拟电路故障特征提取方法[J].电工技术学报,2018,33(11):158-165.Yuan Lifen,Sun Yesheng,He Yigang,et al.Fault feature extraction method for analog circuit based on preferred wavelet packet[J].Transactions of China Electrotechnical Society,2018,33(11):158-165(in Chinese).
[7]Gao Z W,Cecati C,Ding Seven X.A survey of fault diagnosis and fault-tolerant techniques-Part I:Fault diagnosis with model-based and signal-based approaches[J].IEEE Transactions on Industrial Electronics,2015,62(6):3757-3767.
[8]Wang D,Peter W T.Prognostics of slurry pumps based on a moving-wear degradation index and a general sequential Monte Carlo method[J].Mechanical Systems and Signal Processing,2015,56-57:213-229.
[9]周国亮,朱永利,王桂兰.实时大数据处理技术在状态监测领域中的应用[J].电工技术学报,2014,29(S1):432-437.Zhou Guoliang,Zhu Yongli,Wang Guilan.Real-time big data processing technology application in the field of state monitoring[J].Transactions of China Electrotechnical Society,2014,29(S1):432-437(in Chinese).
[10]朱永利,李莉,宋亚奇,等.ODPS平台下的电力设备监测大数据存储与并行处理方法[J].电工技术学报,2017,32(9):199-210.Zhu Yongli,Li Li,Song Yaqi,et al.Storage and parallel processing of big data of power equipment condition monitoring on ODPS platform[J].Transactions of China Electrotechnical Society,2017,32(9):199-210(in Chinese).
[11]AYDMJ T,Duin R P W.Pump failure determination using support vector data description[C]//Proceedings of the19th ACM Conference on Information and Knowledge Management,Toronto,2010.
[12]梅飞,梅军,郑建勇,等.基于KPCA-SVM的断路器故障稳健诊断方法[J].电工技术学报,2014,29(S1):50-58.Mei Fei,Mei Jun,Zheng Jianyong,et al.The robust fault diagnostic method for circuit breaker based on KPCA and SVM[J].Transactions of China Electrotechnical Society,2014,29(S1):50-58(in Chinese).
[13]廖剑,史俊贤,周绍磊,等.基于局部图嵌入加权罚SVM的模拟电路故障诊断方法[J].电工技术学报,2016,31(4):28-35.Liao Jian,Shi Junxian,Zhou Shaolei,et al.Analog circuit fault diagnosis based on local graph embedding weightedpenalty SVM[J].Transactions of China Electrotechnical Society,2016,31(4):28-35(in Chinese).
[14]孙曙光,张强,杜太行,等.基于振动信号的低压万能式断路器分合闸故障程度评估方法的研究[J].中国电机工程学报,2017,37(18):5473-5482.Sun Shuguang,Zhang Qiang,Du Taihang,et al.Study of evaluation method for low voltage conventional circuit breaker switching fault degree based on vibration signal[J].Proceedings of the CSEE,2017,37(18):5473-5482(in Chinese).
[15]史丽萍,汤家升,王攀攀,等.采用最优小波树和改进BP神经网络的感应电动机定子故障诊断[J].电工技术学报,2015,30(24):38-45.Shi Liping,Tang Jiasheng,Wang Panpan,et al.Stator fault diagnosis of induction motors using the optimal wavelet tree and improved BP neural network[J].Transactions of China Electrotechnical Society,2015,30(24):38-45(in Chinese).
[16]杨凌霄,朱亚丽.基于概率神经网络的高压断路器故障诊断[J].电力系统保护与控制,2015,43(10):62-67.Yang Lingxiao,Zhu Yali.High voltage circuit breaker fault diagnosis of probabilistic neural network[J].Power System Protection and Control,2015,43(10):62-67(in Chinese).
[17]Shao H,Jiang H,Zhang X,et al.Rolling bearing fault diagnosis using an optimization deep belief network[J].Measurement Science and Technology,2015,26(11):11502.
[18]Wen L,Gao L,Li X Y.A new deep transfer learning based on sparse auto-encoder for fault diagnosis[J].IEEETransactions on Systems Man and Cybernetics:Systems,2017(99):1-9.
[19]候文擎,叶鸣,李巍华.基于改进堆叠降噪自编码的滚动轴承故障分类[J].机械工程学报,2018,54(7):87-96.Hou Wenqing,Ye Ming,Li Weihua.Rolling element bearing fault classification using improved stacked de-noising auto-encoders[J].Journal of Mechanical Engineering,2018,54(7):87-96(in Chinese).
[20]Ince T,Kiranyaz S,Eren L,et al.Real-time motor fault detection by 1-d convolutional neural networks[J].IEEETransactions on Industrial Electronics,2016,63(11):7067-7075.
[21]Reddi S J,Kale S,Kumar S.On the convergence of adam and beyond[C]//International Conference on Learning Representations(ICLR),Vancouver,Canada,2018.