在很多实际场合下，平稳性的假设往往不能成立，此时就需要引入非平稳信号或时变信号的分析与处理。所谓的时变是指信号的统计特性随时间变化。由于信号的这种性质，传统的傅立叶分析方法无法令人满意。对于时变信号，了解不同时刻附近的频域特征是至关重要的，因此，人们往往采用时间－频率联合描述的方法描述时变信号。-In many practical situations, stationarity assumptions often can not be established, then we need to introduce the analysis and treatment of non-stationary signals or time-varying signal. Refers to the so-called time-varying statistical characteristics of the signal varies over time. Because of the nature of this signal, the conventional method of Fourier analysis is not satisfactory. For time-varying signal, the frequency domain characteristics of different understanding of time is close to critical, therefore, people tend to use time- varying signal frequency method described in conjunction with the descr iption.