基于时频联合分布的经验模态分解方法

Empirical Mode Decomposition Method Based on Joint Time-Frequency Distribution

经验模态分解没有固定的先验基底,使信号分析更加灵活多变,是信号处理领域内解决非线性、非平稳信号分析问题的新方法。但其只可分解信号持续时间内由连续的不同频率成分组成的信号,不能分解不同时间段含有相似频率成分的信号。时频联合分布通过设计时间和频率的联合函数,利用时域和频域联合描述信号在不同时间和频率的能量密度。将时频联合分布应用于其中,扩展经验模态分解方法可处理的信号类型。

Without fixed priori base, empirical mode decomposition （EMD） was a new method to solve the nonlinear and non-stationary signal analysis problems in signal processing field. It made signal analysis more flexible. But EMD was only effective to the signal composed by different succession frequency components within its duration, other than the signal composed by similar frequency components in different time periods. Joint time-frequency distribution described the different time and frequency energy densities of signal in both time and frequency domain by designed joint time-frequency functions. EMD with joint time- frequency could expend the suitable signal type of the original method.