基于希尔伯特谱的瞬时频率滤波方法

A Method of Instantaneous Frequency Domain Filtering Based on HHT

含有色噪声的非平稳信号的滤波去噪一直是未能很好解决的问题,传统的傅立叶分析方法不适用于此类信号的分解及滤波,而小波域滤波效果也不理想。本文提出一种基于希尔伯特谱的瞬时频率滤波方法,并和经验模态分解滤波、基于经验模态分解的小波阀值滤波方法的信噪比进行比较。仿真实验结果表明基于希尔伯特谱的信号分解及滤波去噪方法具有更高的信噪比。

Classical fourier anlaysis can not be used in the filtering or denoising of non-stationary signal, and wavelet approach does not perform well enough in this application. The paper proposed a filtering method based on hilbert spectrum and instantaneous frequency. By contrast with EMD filter and wavelet threshold filter based on EMD, the experimental results the extensive applicability and efficient of this method.