Based on the normal equation algorithm to fir system identification and wavelet iteration , a method to estimate the wavelet from the third - order cumulant of field data is developed . due to the fact that higher order cumulant retains the phase information of the signal , and can suppress the gaussian noise ( color or white noise ) naturally , we can improve the time resolution via inverse filtering the wavelet estimated 本文在基于高阶累积量的fir系统辨识基础上,利用观测信号的三阶累积量对探地雷达子波进行估计,提出一种简单的迭代算法改善了波估计性能,据此进行反褶积,改善反射信号的信杂比和时间分辨率。
It is demonstrated that the solution for 141 000 order normal equations for a simulated astro - geodetic network ( 47 057 points ) and a space one ( 476 points ) iterates 4 993 times to converge to 1 10 - 18 second , taking 28 min and 47 seconds , on a pii / 233 computer , showing over 40 times more efficiency , as compared with the coefficients unadjusted 利用模拟的天文大地网( 47057点)和空间网( 476点)在pii / 233微机上解算14 . 1万阶方程组共迭代4993次(收敛至1 10 - 18角秒) ,花费机时28分47秒,与未进行系数调整相比,功效提高40多倍。