Performance Comparison of Dense Frequency Identifications for Empirical Mode Decomposition
Empirical Mode Decomposition(EMD) suffers some difficulties in separating dense frequencies. The Wavelet Packet Transform (WPT) and Singular-Value Decomposition (SVD) as signal preprocessors were used to decompose a simulated signal with dense frequency components and the performances of two signal preprocess technologies were compared in this paper. The results show that Singular-Value Decomposition (SVD) as preprocessor was better in separating dense frequencies than Wavelet Packet Transform (WPT).
Brendan Gan, Yu Gan and Y. Yu
M. Zheng and F. Shen, "Performance Comparison of Dense Frequency Identifications for Empirical Mode Decomposition", Advanced Materials Research, Vols. 378-379, pp. 266-269, 2012