太阳黑子数时间序列的奇异谱分析和小波分析

Singular spectrum analysis and wavelet analysis on time series of Sunspot

本文对小波变换和奇异谱分析方法进行了简要介绍,对离散小波的分解和重构、奇异谱分析的重构进行了详细阐述。结合太阳黑子数1749年至2007年3月期间的月平均值时间序列进行了小波变换的分解和重构及SSA方法的重构,提取了其主要的周期特性,并对两种分析方法进行了比较。

In this paper,the wavelet transform and singular spectrum analysis(SSA) are introduced briefly.The decomposition and reconstruction of discrete wavelet as well as reconstruction of SSA are expatiated.The SSA and discrete wavelet are then used to analyze the monthly mean of sunspot from 1749 to 2007.Using reconstruction,the original time series can be reconstructed with noise compressed.From the wavelet and SSA analysis,one can obtain the main characteristics of the original in both time and frequency domain.Also,the characteristics derived from the two method are compared.