摘要
经验模态分解由于采用了三次样条插值的方法筛分内模函数,在实际应用中存在模式混叠、在低频段产生多余IMF分量等问题。文中以内模函数定义为出发点,提出一种基于小波包的筛分方法,并通过设定以小波能量比为条件的门限值,提高了小波包分解的自适应性和效率。通过仿真信号和滚动轴承故障振动信号的检验,证明了该方法的可行性和有效性。
Due to applying cubic spline interpolation to sift intrinsic mode functions （IMF） from signal, empirical mode decomposition may generate a series of problem as the phenomena of mode mixing and undesirable IMFs at low frequency region. Here, a sifting method based on wavelet packet was introduced, Starting with definition of intrinsic mode function, through setting the threshold on wavelet energy proportion, it increased the adaptability and efficiency of wavelet packet decomposition. After testing by simulated signals and the rolling bearing fault vibration signal, the feasible and effective of the method proved.
出处
《微计算机信息》
北大核心
2008年第34期158-159,157,共3页
Control & Automation
关键词
经验模态分解
小波包分解
内模函数
小波能量
Empirical Mode Decomposition
Wavelet Packet
Intrinsic Mode Function
Wavelet Energy
