基于EMD和盒维数的Wigner-Ville分布交叉项的抑制方法

Suppression Methods for Cross-term of Wigner-Ville Distribution Based on EMD and Box Dimension

Wigner-Ville分布能够反映信号能量随时间和频率的变化趋势,但该分布易受到交叉项的干扰,限制了该方法的应用。通过分析复频信号Wigner-Ville分布的数学期望和Wigner-Ville谱图的特点,提出用均值阈值来抑制与真实频率不同的交叉项,用经验模态分解(EMD)和盒维数来抑制与真实频率相同的交叉项。仿真分析表明该方法较好地抑制了Wigner-Ville分布中的交叉项,同时该方法还具有良好的抗噪性。

Wigner-Ville distribution can reflect the signal energy trends over time and frequency, but it is vulnerable to suffer cross-term interference, which limits the application of the method. By analyzing the mathematical expectation of the complex-frequency signal Wigner-Ville distribution and Wigner-Ville spectral characteristics, the paper proposes the mean threshold to suppress cross-term which is false frequency and the empirical mode decomposition(EMD) and box dimension to suppress cross-term which is true frequency.Simulation results show that the method can better suppress the cross-term in Wigner-Ville distribution, while the method also has good noise immunity.