基于EMD和Gabor变换的发动机曲轴轴承故障特征提取
|
摘要
针对发动机振动信号的非平稳特点,提出了一种基于经验模态分解(EMD)和Gabor变换相结合的曲轴轴承故障特征提取新方法。通过EMD方法将发动机非稳态加速振动信号分解成多个本征模态函数(IMF),对与原信号相关性强的前4阶IMF分量进行Gabor变换,从各阶分量Gabor时频分布图的频带能量累加曲线中提取能够反映曲轴轴承磨损故障的频带能量作为故障特征参数。试验结果表明,该方法提取的故障特征参数能敏感地反映曲轴轴承的磨损状态,可作为诊断曲轴轴承故障的重要特征量。
In view of the instability feature of engine vibration signals,a method based on the combination of empirical mode decomposition( EMD) and Gabor transform is proposed to extract the fault features of crankshaft bearing. Firstly by using EMD technique the unstable acceleration vibration signals of engine are decomposed into a series of intrinsic mode functions( IMFs). Then Gabor transform is performed on the first 4 orders of IMF components having strong correlation with origin signals. Finally the frequency band energy,which well reflects the wear fault of crankshaft bearing,is extracted as fault characteristic parameter from the frequency band energy accumulation curve of Gabor time / frequency distribution graph for each IMF component. The test results indicate that the fault characteristic parameter extracted with the method proposed can sensitively reflect the wear states of crankshaft bearing and can be taken as the important characteristic quantity for the diagnosis of crankshaft bearing faults.
In view of the instability feature of engine vibration signals,a method based on the combination of empirical mode decomposition( EMD) and Gabor transform is proposed to extract the fault features of crankshaft bearing. Firstly by using EMD technique the unstable acceleration vibration signals of engine are decomposed into a series of intrinsic mode functions( IMFs). Then Gabor transform is performed on the first 4 orders of IMF components having strong correlation with origin signals. Finally the frequency band energy,which well reflects the wear fault of crankshaft bearing,is extracted as fault characteristic parameter from the frequency band energy accumulation curve of Gabor time / frequency distribution graph for each IMF component. The test results indicate that the fault characteristic parameter extracted with the method proposed can sensitively reflect the wear states of crankshaft bearing and can be taken as the important characteristic quantity for the diagnosis of crankshaft bearing faults.
引文
[1]肖云魁.汽车故障诊断学[M].北京:北京理工大学出版社,2006.
[2]赵晓平,张令弥,郭勤涛.旋转机械阶比跟踪技术研究进展综述[J].地震工程与工程振动,2008,28(6):213-214.
[3]Huang Norden E,Shen Zheng,Long Steven R,et al.The Empirical Mode Decomposition and the Hilbert Spectrum for Nonlinear and Non-stationary Time Series Analysis[J].Proc.R.Soc.Lond.A,1998,454:903-995.
[4]张超,陈建军,徐亚兰.基于EMD分解和奇异值差分谱理论的轴承故障诊断方法[J].振动工程学报,2011,24(5):539-545.
[5]吴虎胜,吕建新,战仁军,等.基于EMD和AR奇异值的柴油机故障诊断[J].机械设计与制造,2011,4(4):230-232.
[6]夏天,王新晴,肖云魁,等.应用EMD-AR谱提取柴油机曲轴轴承故障特征[J].振动、测试与诊断,2010,30(3):318-321.
[7]张玲玲,廖红云,曹亚娟,等.基于EEMD和模糊C均值聚类算法诊断发动机曲轴轴承故障[J].内燃机学报,2011,29(4):332-336.
[8]李辉,郑海起,唐力伟.基于EMD和功率谱的齿轮故障诊断研究[J].振动与冲击,2006,25(1):133-136.
[9]蔡艳萍,李艾华,王涛,等.基于EMD-Wigner-Ville的内燃机振动时频分析[J].振动工程学报,2010,23(4):430-437.
[2]赵晓平,张令弥,郭勤涛.旋转机械阶比跟踪技术研究进展综述[J].地震工程与工程振动,2008,28(6):213-214.
[3]Huang Norden E,Shen Zheng,Long Steven R,et al.The Empirical Mode Decomposition and the Hilbert Spectrum for Nonlinear and Non-stationary Time Series Analysis[J].Proc.R.Soc.Lond.A,1998,454:903-995.
[4]张超,陈建军,徐亚兰.基于EMD分解和奇异值差分谱理论的轴承故障诊断方法[J].振动工程学报,2011,24(5):539-545.
[5]吴虎胜,吕建新,战仁军,等.基于EMD和AR奇异值的柴油机故障诊断[J].机械设计与制造,2011,4(4):230-232.
[6]夏天,王新晴,肖云魁,等.应用EMD-AR谱提取柴油机曲轴轴承故障特征[J].振动、测试与诊断,2010,30(3):318-321.
[7]张玲玲,廖红云,曹亚娟,等.基于EEMD和模糊C均值聚类算法诊断发动机曲轴轴承故障[J].内燃机学报,2011,29(4):332-336.
[8]李辉,郑海起,唐力伟.基于EMD和功率谱的齿轮故障诊断研究[J].振动与冲击,2006,25(1):133-136.
[9]蔡艳萍,李艾华,王涛,等.基于EMD-Wigner-Ville的内燃机振动时频分析[J].振动工程学报,2010,23(4):430-437.
版权所有：© 2023 中国地质图书馆 中国地质调查局地学文献中心