一种新的地震子波估计方法

A New Method for Seismic Wavelet Estimation

地震子波估计问题是地震勘探信号处理和分析中的关键一环。本文基于遗传算法提出了一个新的地震子波估计方法。该方法用ARMA模型描述地震子波，用遗传算法交替迭代地估计AR和MA参数。与其它方法相比，本文提出的方法具有高度稳定性和很好的精度，并适应于估计非最小相位地震子波。

The estimation of seismic wavelet is an important problem in the seismic signal processing. In this paper we proposed a new method for seismic wavelet estimation based on the genetic algorithms. We use ARMA model to describe the seismic wavelet, and use genetic algorithms to estimate the AR parameters and the MA parameters iteratively. Comparing with current method, it has high stability, good precision and can be use to estimate non minimum seismic wavelet.