Abstract：Thin-bed thickness computation using amplitude tuning is very subject to the influence of data quality,so the accuracy is not high.The method is accurate and noise-resistant which derives thin bed thickness from the linear or simple nonlinear expression that was formed according to the relation between the thickness and dominant frequency dominant amplitude;however,the method is also subject to the influence of seismic data quality.Neural network BP modeling method estimates desirable thin bed thickness because it takes known borehole data as learning sample to automatically model the relation between thin bed thickness and amplitude.frequency.In addition,the combinative method,according to which we estimate the thin bed thickness by first deriving the.linear fitted thickness from amplitude and lowfrequency energy percentage then by computing residual error with the use of BP modeling,brings more satisfactory result than the only lip modeling does.
龙建东. 薄层厚度与振幅、频率关系的神经网络BP法模拟[J]. 石油地球物理勘探, 1995, 30(6): 817-822.
Long Jianong. Neural network BP modeling of the relation between thin bed thickness and amplitude&frequency. OGP, 1995, 30(6): 817-822.
