用时频分析方法分析非平稳信号

Analysis of non-stational signals by time-frequency analysis methods

目前经典信号处理技术，即时域分析和频域分析，尚不能很好地适应许多科学领域中普遍存在着的非平稳信号的分析，我们把已发展起来的三种时频分析方法：短时傅里叶变换、Ｗｉｇｎｅｒ分布及小波变换，应用到非平稳信号分析中，对这三种方法的原理、性质、优缺点及实际应用逐一进行了论述。

It is in urgent need to analyze non- stationaly signals existing extensively in the fields of many sciences for extracting more characteristic infor-mation、 whereas. at present the classical signal processing methods,i.e, time-do-main analysis and frequency-domain analysis,unfit for analyzing this kind of signals. We apply three recently developed time-frequency analysis methods:they are short-time Fourier transformation,Wigner distribution and wavelet trans-form,analysis of non-stationary signals.The concepts,principles,properties,ad-vantages and disadvantages of these three methods are discussed respectively,and some typical applications are performed to prove their feasibilities.