局域能量逐层提取的LMD模态混叠抑制方法

提出了一种基于局域能量逐层提取的局域均值分解（Local Mean Decomposition，LMD）模态混叠抑制方法。对待分析信号利用LMD得到一系列乘积函数（Product Function，PF），对各PF的能量进行积分，根据积分比值确定模态混叠导致的能量泄漏程度，并从具有模态混叠的高阶PF中移除下一阶PF的成分构成新的PF，从原始信号中减去新的PF得到新的待处理信号，继续执行LMD，直到所有PF频率成分各自独立，从而实现对模态混叠的抑制。仿真结果表明：算法可以有效抑制小频率比混合信号导致的模态混叠，在实际工程上具有一定推广应用价值。

LMD Mode Mixing Suppression Method by Layer Local Energy Extraction

An improved algorithm based on local energy layer by local mean decomposition (LMD)layer extraction was proposed. The signal was decomposed to series Product Function (PF)by LMD and the local energy of each PF was obtained by integration. The leakage of energy level can be determined according to the local energy ratio,based on which the mode mixing component can be removed from the high order PF. The new PF was subtracted from the original signal and LMD was carried on repeatedly untill all the PF was obtained. The improved LMD can ensure that all PF frequency components are independent of each other,so as to realize the suppression of mode mixing. The simulation results show that the algorithm can effectively overcome the mode mixing caused by near frequency signal in the multicomponent signal,and it has certain application value in practical engineering.