基于Hilbert-Huang变换的高斯线性调频信号的时频分析

Based on the Empirical Mode Decomposition algorithm, Hilbert-Huang transform (HHT) is a data-driven and self-adapting method of energy-time-frequency distribution and is able to compose any complicated data into Intrinsic Mode Functions, which has meaningful instantaneous frequency in physical and limited errors of residues. This paper quantifies and describes the numerical relationship between the EMD error and the sampling frequency using single component cosine signal for example. The result reveals that increasing the sampling frequency will improve the frequency resolution. We also analyze the time-frequency assemble quality of HHT of Gaussian Linear Frequency Modulation signal.