摘要
本文介绍的矢量波场分离方法可实现对多分量非零井源距VSP资料的波场分离,该方法是一种局部参数反演方法,假定单道记录由上、下行纵波和上、下行横波叠加而成,每种波由傅里叶分量(波形)、视慢度和偏振角度构成。再利用邻近几道记录在有效频带内建立该反演问题的正演模型。当对多分量VSP资料进行波场分离时,往往要在正演模型中引入偏振矢量,这样就增加了反演问题解的惟一性和稳定性,但带来了非线性化问题。本文利用施加约束条件的优化方法,快速有效地计算出最优的偏振角度、速度值,克服了这种非线性问题,并通过求解此线性方程组得出各个波的傅里叶分量信息,最终由逆傅里叶变换到时间域实现了波场分离,也附带求出了各种波(上、下行纵波和上、下行横波)的速度和偏振角度等参数。经过对多分量非零井源距VSP模型数据及实际资料进行波场分离处理,验证了该方法的稳定性和实用性。
In this paper the authors introduced the vector wave field separation method which can realize the wave field separation for multi-component non-zero offset VSP data,it is a regional parameter inversion method in which the single channel record is the stack of down-going and up-going P-wave and down-going and up-going share wave,and each wave consists of Fourier component (wave form),apparent slowness and polarization angle. By utilizing the nearby traces the forward model of the inversion problem was established. When separating wave field for multi-component VSP data,polarization vector was generally brought in the forward model so that the uniqueness and stability of the resolution of the inversion problem got raised,but non-linear problem was brought in. In this paper by utilizing the optimization algorithm which constraint conditions were applied,the optimal polarization angle and velocity values were calculated and the non-linear problem was overcame,the linear equations were solved so the Fourier component information for each wave were obtained,at last wave separation was realized by inverse Fourier transform in time domain,at the same time the velocity for the every type of wave (down-going and up-going P-wave and down-going and up-going share wave) ,polarization angle and other parameters were calculated,the stability and practicability of this method was proven by wave field separation processing for the multi-component non-zero VSP model data and field data.
出处
《石油地球物理勘探》
EI
CSCD
北大核心
2009年第6期708-719,共12页
Oil Geophysical Prospecting
基金
国家重点基础研究计划“973”项目“中国西部典型叠合盆地油气成藏机制与分布规律”(2006CB202300)
关键词
VSP
矢量波场分离
参数反演
视慢度
偏振角度
优化算法
VSP vector wave field separation,parameter inversion,apparent slowness,polarization angle,optimization algorithm,constraint conditions