Research on EMD improvement methods

Abstract

The article focuses on the basic concept and theoretical basis of EMD, and analyzes the advantageous aspects and shortcomings of EMD. It also introduces various methods to improve EMD, such as the proposed EEMD method, which describes the theory and advantages of EEMD, the EEMD-CNN method, which is to eliminate the added Gaussian white noise first, before the decomposition of EEMD to improve its accuracy, the CEEMD processing, which adds two opposite Gaussian white noises, and the MEEMD processing, in which the CEEMD processing adds the concept of cliff value to improve its decomposition efficiency and make it more practical.