匹配 Wigner 变换及其在瞬时频率估计中的应用

摘要

A novel time-frequency analysis method referred to as matched Wigner transform (MWT) was defined in this pa-per .By matching bilinear signal kernel with Taylor extended Fourier revolving factor ,MWT can produce a time-frequency distribu-tion (TFD ) with best concentration and without inner artifacts for mono-component high-order polynomial phase signal (PPS ) .In addition ,an effective parameter and instantaneous frequency (IF ) estimation algorithm was proposed based on MWT .The algorithm was validated by simulation examples .Finally ,excellent sea experiment results were obtained when applying it to analyze the weak Doppler signal of underwater moving acoustic source .%提出了一种新的时频分析方法，称为匹配Wigner变换（Matched Wigner Transform ，MWT ）。该变换通过泰勒级数扩展傅立叶旋转因子e－jωτ来匹配双线性信号核，能够对单分量高阶多项式相位信号（Polynomial Phase Signal ， PPS ）分析得到聚集性最优的时频分布（Time Frequency Distribution ，TFD ），且克服了自交叉项的干扰。基于MWT ，给出了一种有效实现PPS的瞬时频率（Instantaneous Frequency ，IF）和参量估计的迭代算法，并通过仿真算例验证了算法的有效性。最后，将该方法应用于水下运动声源的微弱多普勒信号分析，得到良好的海试试验结果。