摘要
针对传统经验模态分解(EMD)方法中信号所含频率分量较接近易出现模态混叠现象,导致分解出的本征模态函数失去原有物理意义问题,提出改进算法。将原始信号频带信号微分;对微分信号进行EMD分解;将所得IMFs进行积分处理获得原始信号有效的IMFs,能抑制EMD分解过程中的模态混叠现象。仿真计算与实际工程应用表明,该方法优于传统EMD算法,可有效分离频率接近的本征模态函数,扩大EMD方法的适用范围。
When a signal contains components with frequencies close to each other a mode smixing phenomenon occurs, using the traditional empirical mode decomposition （EMD） method to process it the decomposed intrinsic mode functions lose their physical meaning. Aiming at this phenomenon, a new processing method was presented. A differential operator was used to differentiate the original signal in time domain, then the differentiated signal was decomposed with EMD, the obtained IMFs were integrated to get effective IMFs of the original signal. The results showed that the proposed method can suppress modal aliasing. Simulations and practical engineering applications indicated that this method is superior to the traditional EMD; it can not only effectively separate intrinsic mode functions with frequencies close to each other, but also expand the application scope of EMD.
出处
《振动与冲击》
EI
CSCD
北大核心
2013年第18期133-135,145,共4页
Journal of Vibration and Shock
关键词
振动与波
经验模态分解
频带分解
vibration and wave
empirical mode decomposition （EMD）
frequency-band separating
