摘要
针对Hilbert-Huang变换(HHT)的不足,应用时间序列重构矩阵奇异值分解的能量分析方法改进了HHT,并提出了用来识别伪固有模态函数(IMF)的方法.利用仿真信号检验了改进HHT的效果.将改进后的HHT应用到某双转子航空发动机轻碰摩故障仿真试验结果的检测中,并与改进前HHT和小波分析的结果进行了比较,得出了改进后的HHT具有有效分离出微弱信号分量等优势,且能比小波变换更好地确定碰摩的存在.
In order to overcome the shortages of Hilbert-Huang transform（HHT）,an improved HHT was proposed by the energy analysis method based on the singularity value decomposition of the track matrix reconstructed by time series,and a method for judging the pseudo-IMFs by empirical mode decomposition（EMD） was also presented.The advantages of improved HHT were verified from simulation signals.Then,the improved HHT was applied to detect the slight rubbing simulation of dual-rotor aero engine.The simulation results were compared with previous HHT and wavelet transform,respectively,showing that the improved HHT has the advantage of separating efficiently the weak component of signals from other strong components,and also confirming the slight rubbing more easily.The analytical results provide a useful reference for diagnosing the fault of rubbing.
出处
《航空动力学报》
EI
CAS
CSCD
北大核心
2007年第10期1779-1784,共6页
Journal of Aerospace Power
基金
航空科学基金(03I51037)
关键词
航空、航天推进系统
改进HHT
碰摩
经验模态分解(EMD)
故障诊断
aerospace propulsion system
improved Hilbert-Huang transform（HHT）
rub
empirical mode decomposition（EMD）
fault diagnose
作者简介
何田（1979-），男，四川广安人，博士生，主要从事机械系统动力学及故障诊断的研究．
