Outlier detection based on Hilbert-Huang Transform

In this paper, we review the main concept and the developing method of Hilbert-Huang Transform (HHT), which is the novel and high-efficiently nonlinear and non-stationary data analysis method. HHT is composed of Empirical Mode Decomposition(EMD) and Hilbert Spectral Analysis(HSA). Numerical experiments demonstrate the advantages of HHT at outliers detection for inter-wave modulation signal compared with wavelet decomposition. Furthermore, based on the HHT and EEMD, a new frequency analysis method for intra-wave modulation signal is proposed. Simulation shows that the proposed method tracks the frequency of intra-wave modulation signal precisely.