基于加窗四阶累量时延估计的微瞬态电磁辐射信号检测算法

以研究微弱瞬态电磁辐射信号的探测手段为目的,提出了一种基于加窗四阶累量时延估计的检测算法。利用高阶累量对高斯噪声的抑制作用以及相干平滑变换对微弱窄带噪声的弱化效果,将低信噪比条件下非周期微弱放电信号的检测问题转化为周期性时延参数的估计问题。仿真及实验比较了加窗四阶累量互相关、非加窗四阶累量互相关及传统互相关时延估计3种算法的检验性能。结果表明,经过加窗处理的四阶累量时延估计算法在相关高斯白噪声、非相关高斯白噪声及窄带噪声3种噪声条件下的检测性能都优于其他两种方法,比较适合用于微弱电磁辐射源的远距离探测。

Weak transient electromagnetic signal detection based on windowing forth-order cumulants and cross correlation time-delay estimation

To investigate the remote detecting approaches of transient electromagnetic signals, the detecting method based on windowing forth order cumulants and cross correlation time delay estimation algorithm is proposed. For the interference from Gaussian noise and narrow band noise, smoothed coherence transform (SCOT) forth order cumulants of signal and the cross correlation algorithm for time delay estimation are introduced. The signal detection of weak non periodic discharge is translated to the estimation of periodic time delay parameters. The detecting performances of the three methods, windowing forth order cumulants, forth order cumulants and conditional cross correlation, are analyzed by simulation and experiments. The results show that the windowing forth order cumulants time delay estimation method is superior to the other two approaches in the three noise conditions, so the pulse accumulation is suitable for remote detection of partial discharge source.