【摘要】 基于HHT(Hilbert-Huang transformation)是一种能分解出信号的任何频率分量的主成分分析法以及线性调频信号(LFM)的瞬时频率是关于时间的直线,将LFM信号先作HHT得到其瞬时频率,然后利用该瞬时频率中间部分的时频点作最小二乘(LSM)直线拟合。这种改进的LSM直线拟合方法不仅剔除了HHT固有的边界效应在瞬时频率斜线两端产生的高频谐波,从而有效地抑制了这种边界效应对瞬时频率估计的影响,而且直接得到了较为准确的LFM信号的瞬时频率参数估计值。与通过改变HHT算法来减小HHT边界效应的方法相比,该方法更有效、更简单易行。仿真结果表明,该方法较小波脊频率提取法更具优越性。更多还原

【Abstract】 According to the fact that Hilbert-Huang transformation(HHT) is a main component analysis method and that linear frequency modulation(LFM) signals instantaneous frequencies are lines with respect to variation time.This paper combined the HHT and least squares method to implement their instantaneous frequency estimations.Moreover only used the middle of instantaneous time-frequency samples obtained by HHT to fit instantaneous frequency line,which not only could reduce the influence of the end effects inherent in HHT on instantaneous frequency estimation more simply and efficiently than that changing the HHT algorithm did,but also the much more precision instantaneous parameter estimations could be directly achieved.Simulation results show that the proposed method has many advantages over the wavelet ridge method in term of instantaneous frequency estimation of LFM signals.更多还原