运用经验模态分解（EMD）、集合经验模态分解（EEMD）和完备总体经验模态分解（CEEMD）3种方法对原始基线时间序列进行分解，得到各自时间序列的本征函数模态分量（IMF）、相关系数及周期性强度，进而确定其季节项分量，同时通过比较季节项和原始基线时间序列的叠加功率谱图优选分解方法。结果表明，CEEMD方法对基线时间序列季节项提取和重构效果最佳。"/>
In this paper, empirical mode decomposition(EMD), ensemble empirical mode decomposition(EEMD) and complete ensemble empirical mode decomposition(CEEMD) are used to decompose the original baseline time series. The corresponding time series intrinsic mode function(IMF), correlation coefficients and periodic intensity are obtained, and the season components are determined. The decomposition method is optimized by comparing the superposition power spectrum of the season item with the original baseline time series. The results show that CEEMD method has the best effect on extracting and reconstructing season item of baseline time series."/>
Analysis of Season Item Extraction and Reconstruction of Long Baseline Time Series
Abstract In this paper, empirical mode decomposition(EMD), ensemble empirical mode decomposition(EEMD) and complete ensemble empirical mode decomposition(CEEMD) are used to decompose the original baseline time series. The corresponding time series intrinsic mode function(IMF), correlation coefficients and periodic intensity are obtained, and the season components are determined. The decomposition method is optimized by comparing the superposition power spectrum of the season item with the original baseline time series. The results show that CEEMD method has the best effect on extracting and reconstructing season item of baseline time series.
