基于变分模态分解和符号熵的齿轮故障诊断方法
|
摘要
为提高齿轮的故障诊断效果,提出了基于变分模态分解(Variational Modal Decomposition,VMD)和符号熵(Symbol Entropy, SE)的齿轮故障诊断方法。首先,利用VMD对齿轮故障振动信号进行分解,得到若干个本征模态分量(Intrinsic Mode Function,IMF);然后,计算IMF分量的符号熵,并将IMF符号熵组成齿轮故障特征向量;最后,将特征向量输入SVM进行故障诊断。齿轮故障诊断实测结果验证了该方法的有效性和优势。
In order to improve diagnosis accuracy of gear,a fault diagnosis method of gear based on varia-tional modal decomposition(VMD)and symbol entropy(SE)is proposed. Firstly,the gear vibration signals isdecomposed into several intrinsic mode Function(IMF)which with different frequency components. Secondly,SE values of each IMF are calculated as fault feature vectors that could represent the operating conditions ofgear. Finally,the fault feature is put into SVM to identify different faults. Experiment results of gear show thatthe proposed method has certain superiority.
In order to improve diagnosis accuracy of gear,a fault diagnosis method of gear based on varia-tional modal decomposition(VMD)and symbol entropy(SE)is proposed. Firstly,the gear vibration signals isdecomposed into several intrinsic mode Function(IMF)which with different frequency components. Secondly,SE values of each IMF are calculated as fault feature vectors that could represent the operating conditions ofgear. Finally,the fault feature is put into SVM to identify different faults. Experiment results of gear show thatthe proposed method has certain superiority.
引文
[1]林近山,陈前.基于多重分形去趋势波动分析的齿轮箱故障特征提取方法[J].振动与冲击,2013,32(2):97-101.
[2]刘春林,潘宏侠,史斐娜,等.基于MF-DFA和SVM的齿轮箱故障诊断[J].机械传动,2014,38(12):119-123.
[3]林近山,窦春红,寇兴磊.基于时变滤波经验模式分解的齿轮箱故障诊断[J].机械传动,2018,42(1):98-101.
[4]费春红,赵光胜,寇兴磊.基于CEEMDAN能量熵的齿轮状态识别[J].机械传动,2018,42(1):102-105.
[5]杨望灿,张培林,吴定海,等.基于EEMD和奇异熵增量谱的齿轮故障特征提取[J].机械传动,2014,38(2):141-146.
[6]DRYAGOMIRETSKI K,ZOSSO D.Variational mode decomposition[J].TEEE Transaction on Signal Processing,2014,62(3):531-544.
[7]赵洪山,郭双伟,高夺.基于奇异值分解和变分模态分解的轴承故障特征提取[J].振动与冲击,2016,35(22):183-188.
[8]WANG Y,MARKERT R,XIANG J,etal.Reasrch on variational mode decomposition and its application in detceting rub-impact fault of the rotor sstem[J].Mechanical System and Signal Process,2015,60:243-251.
[9]钱林,康敏,傅秀清,等.基于VMD的自适应形体学在轴承故障诊断中的应用[J].振动与冲击,2017,36(3):227-233.
[10]张淑清,孙国秀,李亮,等.基于LMD近似熵和FCM聚类的机械故障诊断研究[J].仪器仪表学报,2013,34(3):714-720.
[11]向丹,葛爽.基于EMD样本熵-LLTSA的故障特征提取方法[J].航空动力学报,2014,29(7):1535-1542.
[12]张良,张前图.基于LCD模糊熵和流行学习的故障特征提取方法[J].机械强度,2016,38(2):225-230.
[13]贾峰,武兵,熊晓燕,等.基于多维度排列熵和支持向量机的轴承早期故障诊断方法[J].计算机集成制造系统,2014,20(9):2275-2282.
[14]DONG Guo Zhu,PEI Jian.Sequence data mining[M].New York,USA:Springer-Verlag New York Inc,2007:31-32.
[15]张豪,陈黎飞,郭躬德.基于符号熵的序列相似性度量方法[J].计算机工程,2016,42(5):201-207.
[16]徐卓飞,刘凯.基于极值符号序列分析的EMD端点效应处理方法[J].振动、测试与诊断,2015,35(2):306-315.
[17]和卫星,陈晓平.LZ复杂度算法中的二值化方法分析及改进[J].江苏大学学报(自然科学版),2004,26(3):261-264.
[2]刘春林,潘宏侠,史斐娜,等.基于MF-DFA和SVM的齿轮箱故障诊断[J].机械传动,2014,38(12):119-123.
[3]林近山,窦春红,寇兴磊.基于时变滤波经验模式分解的齿轮箱故障诊断[J].机械传动,2018,42(1):98-101.
[4]费春红,赵光胜,寇兴磊.基于CEEMDAN能量熵的齿轮状态识别[J].机械传动,2018,42(1):102-105.
[5]杨望灿,张培林,吴定海,等.基于EEMD和奇异熵增量谱的齿轮故障特征提取[J].机械传动,2014,38(2):141-146.
[6]DRYAGOMIRETSKI K,ZOSSO D.Variational mode decomposition[J].TEEE Transaction on Signal Processing,2014,62(3):531-544.
[7]赵洪山,郭双伟,高夺.基于奇异值分解和变分模态分解的轴承故障特征提取[J].振动与冲击,2016,35(22):183-188.
[8]WANG Y,MARKERT R,XIANG J,etal.Reasrch on variational mode decomposition and its application in detceting rub-impact fault of the rotor sstem[J].Mechanical System and Signal Process,2015,60:243-251.
[9]钱林,康敏,傅秀清,等.基于VMD的自适应形体学在轴承故障诊断中的应用[J].振动与冲击,2017,36(3):227-233.
[10]张淑清,孙国秀,李亮,等.基于LMD近似熵和FCM聚类的机械故障诊断研究[J].仪器仪表学报,2013,34(3):714-720.
[11]向丹,葛爽.基于EMD样本熵-LLTSA的故障特征提取方法[J].航空动力学报,2014,29(7):1535-1542.
[12]张良,张前图.基于LCD模糊熵和流行学习的故障特征提取方法[J].机械强度,2016,38(2):225-230.
[13]贾峰,武兵,熊晓燕,等.基于多维度排列熵和支持向量机的轴承早期故障诊断方法[J].计算机集成制造系统,2014,20(9):2275-2282.
[14]DONG Guo Zhu,PEI Jian.Sequence data mining[M].New York,USA:Springer-Verlag New York Inc,2007:31-32.
[15]张豪,陈黎飞,郭躬德.基于符号熵的序列相似性度量方法[J].计算机工程,2016,42(5):201-207.
[16]徐卓飞,刘凯.基于极值符号序列分析的EMD端点效应处理方法[J].振动、测试与诊断,2015,35(2):306-315.
[17]和卫星,陈晓平.LZ复杂度算法中的二值化方法分析及改进[J].江苏大学学报(自然科学版),2004,26(3):261-264.