支持向量机与证据理论融合在电机转子故障诊断中的应用
|
摘要
在电机转子故障诊断中,为了进一步提高诊断方法的自适应性和分类准确性,提出一种支持向量机和证据理论的故障诊断方法.利用小波包分解振动信号和提取特征向量,构造多类支持向量机概率输出.采用改进的D-S证据理论,建立支持向量机与证据理论的诊断模型.实验结果表明:与常规故障诊断方法相比,该故障诊断方法可行,且具有更高的诊断准确率,为电机转子故障诊断研究提供有效的途径.
In the fault diagnosis of motor rotor, in order to further improve the adaptability and classification accuracy of the diagnostic method, this paper presents a motor fault diagnosis method based on support vector machine and evidence theory.Decomposition of vibration signals and extraction of feature vectors using wavelet packets, this method constructs multi-class support vector machine probability output. Using improved D-S evidence theory, the article establishes a diagnostic model of support vector machine and evidence theory. Experiments show that, compared with the conventional fault diagnosis method, the fault diagnosis method is feasible and has higher diagnostic accuracy, which provides an effective way for motor rotor fault diagnosis research.
In the fault diagnosis of motor rotor, in order to further improve the adaptability and classification accuracy of the diagnostic method, this paper presents a motor fault diagnosis method based on support vector machine and evidence theory.Decomposition of vibration signals and extraction of feature vectors using wavelet packets, this method constructs multi-class support vector machine probability output. Using improved D-S evidence theory, the article establishes a diagnostic model of support vector machine and evidence theory. Experiments show that, compared with the conventional fault diagnosis method, the fault diagnosis method is feasible and has higher diagnostic accuracy, which provides an effective way for motor rotor fault diagnosis research.
引文
[1]钟秉林,黄仁.机械故障诊断学[M]. 3版.北京:机械工业出版社, 2007.
[2]岳晓峰,刘书溢.基于FastICA算法的转子故障特征分析[J].制造业自动化,2015,37(20):82-86.
[3]唐贵基,庞彬.改进的HVD方法在转子故障诊断中的应用[J].电机与控制应用,2015,42(2):73-78.
[4]张维强,赵荣珍,李坤杰,等.基于MSKPCA和SVM的转子故障诊断模型及应用[J].机械设计与制造,2015(10):4-8.
[5]苟旭丹.基于Hilbert模量与改进BP神经网络的电机转子断条故障诊断[J].电测与仪表,2018,55(3):55-58.
[6]荆双喜,赵行宇,郭松涛,等.异步电机转子断条故障诊断研究[J].河南理工大学学报(自然科学版),2016,35(2):224-229.
[7]李伟伟,王莉,张琳,等.基于改进LS-SVM的异步电机转子故障诊断[J].火力与指挥控制,2016,41(2):136-141.
[8]石江波,杨兆建,郭伟杰,等.电流与振动信号融合的转子系统故障诊断研究[J].机械设计与制造,2018(8):19-21,25.
[9]李善,谭继文,俞昆.神经网络和改进D-S证据理论相结合的滚动轴承复合故障诊断研究[J].机床与液压,2018,46(1):153-157,184.
[10]夏飞,孟娟,杨平,等.改进D-S证据理论在振动故障诊断中的应用[J].电子测量与仪器学报,2018,32(7):171-179.
[11] CORT ES C, VAPNIK V. Support-vector networks[J]. Machine Learning,1995, 20(3):273-297.
[12]郎国伟,周东方,胡涛,等.基于D-S证据理论的故障诊断方法[J].太赫兹科学与电子信息学报,2017,15(3):465-468.
[13]李亦滔.基于支持向量机和证据理论融合的旋转机械故障诊断研究[D].南昌:华东交通大学,2013.
[2]岳晓峰,刘书溢.基于FastICA算法的转子故障特征分析[J].制造业自动化,2015,37(20):82-86.
[3]唐贵基,庞彬.改进的HVD方法在转子故障诊断中的应用[J].电机与控制应用,2015,42(2):73-78.
[4]张维强,赵荣珍,李坤杰,等.基于MSKPCA和SVM的转子故障诊断模型及应用[J].机械设计与制造,2015(10):4-8.
[5]苟旭丹.基于Hilbert模量与改进BP神经网络的电机转子断条故障诊断[J].电测与仪表,2018,55(3):55-58.
[6]荆双喜,赵行宇,郭松涛,等.异步电机转子断条故障诊断研究[J].河南理工大学学报(自然科学版),2016,35(2):224-229.
[7]李伟伟,王莉,张琳,等.基于改进LS-SVM的异步电机转子故障诊断[J].火力与指挥控制,2016,41(2):136-141.
[8]石江波,杨兆建,郭伟杰,等.电流与振动信号融合的转子系统故障诊断研究[J].机械设计与制造,2018(8):19-21,25.
[9]李善,谭继文,俞昆.神经网络和改进D-S证据理论相结合的滚动轴承复合故障诊断研究[J].机床与液压,2018,46(1):153-157,184.
[10]夏飞,孟娟,杨平,等.改进D-S证据理论在振动故障诊断中的应用[J].电子测量与仪器学报,2018,32(7):171-179.
[11] CORT ES C, VAPNIK V. Support-vector networks[J]. Machine Learning,1995, 20(3):273-297.
[12]郎国伟,周东方,胡涛,等.基于D-S证据理论的故障诊断方法[J].太赫兹科学与电子信息学报,2017,15(3):465-468.
[13]李亦滔.基于支持向量机和证据理论融合的旋转机械故障诊断研究[D].南昌:华东交通大学,2013.