中文说明：EMD分解方法是基于以下假设条件：⑴数据至少有两个极值，一个最大值和一个最小值；⑵数据的局部时域特性是由极值点间的时间尺度唯一确定；⑶如果数据没有极值点但有拐点，则可以通过对数据微分一次或多次求得极值，然后再通过积分来获得分解结果。这种方法的本质是通过数据的特征时间尺度来获得本征波动模式，然后分解数据。这种分解过程可以形象地称之为“筛选（sifting）”过程。本代码包括完整的EMD算法Matlab实现，并含有数据集用法脚本，希望多您有所帮助。
English Description：
The EMD decomposition method is based on the following assumptions: (1) the data has at least two extremums, one maximum and one minimum; (2) the local time-domain characteristics of the data are uniquely determined by the time scale between the extremum points; (3) if the data has no extremum point but inflection point, the extremum can be obtained by one or more differentiation of the data, and then the decomposition result can be obtained by integration. The essence of this method is to obtain the intrinsic wave pattern through the characteristic time scale of the data, and then decompose the data. This decomposition process can be called "sifting" process. This code includes the complete EMD algorithm matlab implementation, and contains a dataset usage script, I hope you can help.