中文说明：应用背景通过经验模式分解(empirical modedecomposition，EMD)将信号分解成一系列单频率信号——本征模函数(intrinsic mode function，IMF),在通过 Hilbert 变换即可求得每个本征模函数的瞬时频率。Hilbert-Huang 变换在实际应用中，存在的最大问题就是端点效应。针对上述问题，提出一种以残差与原信号相关系数为阈值的自适应的虚假 IMF 筛选算法。 关键技术复杂信号分解成单频率信号,任意本征模函数需满足两个条件：(1)在整个数据长度，极值和过零点数目必须相等或最多相差一个；(2)在任意数据点，局部最大值的包络和局部最小值的包络的平均必须为零。
English Description：
Application backgroundBy modedecomposition empirical (EMD), the signal is decomposed into a series of single frequency signal, the intrinsic mode function (mode function; intrinsic, IMF), the instantaneous frequency of each intrinsic mode function is obtained by Hilbert transform. In the practical application of Hilbert-Huang transformation, the biggest problem is the endpoint effect. In order to the above problems, an adaptive false IMF filtering algorithm based on the correlation coefficient between the residual and the original signal is proposed.Key TechnologyThe complex signal is decomposed into a single frequency signal, and arbitrary the intrinsic mode functions need to meet two conditions: (1) in the whole data length, the number of extreme value and zero crossing number must be equal or at most one; (2) in any data point, the average value of local maximum envelope and local minimum value must be zero.