A Contrastive Analysis of EMD and EEMD on Multiple Time-scale Characteristics of Rainfall
Zhang Guo-yon · China Rural Water and Hydropower · 2014
According to the multiple time-scale characteristics analysis of hydrological time series,the mode decomposition is made by empirical mode decomposition(EMD)and ensemble empirical mode decomposition(EEMD)method.In this paper,the EMD is compared to EEMD based on a test function firstly,and which shows the superiority of EEMD in solving the mode mixing problem.And then,the EEMD and EMD method were applied based on least squares support vector machines(LSSVM)signal continuation to analyze the annual rainfall sequence of Hubei Province.Verified by the results of wavelet analysis,as a new signal processing method,the EEMD method,can obtain more accurate characteristics of annual rainfall series than EMD.Based on this,it can be found that the presented method can be widely used for the multi-scale analysis of hydrological time series.