摘要
通过地震和测井联合反演,可以获得高分辨率的井间地层波阻抗分布的信息。综合应用波动方程反演和神经网络分析来反演地层波阻抗参数,其过程分3步:第一步,应用先验地质知识,对地震数据和测井曲线进行地质解释,并对测井曲线进行对层和标定,然后求取相应的层速度的低频信息,旨在搞清井间地层结构状况,为非线性反演提供地层产状的先验信息;第二步,应用非线性波动方程反演,在层速度界面及井中物性参数约束下,从地震数据中反演高分辨率的反射系数及波阻抗参数;第三步,应用CUSI神经网络分析方法,以高分辨率的反射系数及波阻抗等参数作为约束,以沿层求取的地震特征作为输入,以井中反演的波阻抗参数为期望输出,对非线性波动方程反演出的波阻抗参数进行非线性标定,得出井间的地层绝对波阻抗物性参数。
High resolution crosshole distribution of wave impedance can be estimated through seismic and logging joint inversion. The procedure of wave impedance estimation by combined wave equation inversion and neural network analysis consists of three steps. Firstly, to carry out geologic interpretation on seismic and logging data under the priori geologic knowledge, perform correlation and calibration for horizons on logs, and determine low frequency contents of the corresponding interval velocities. Secondly, inversion of high resolution reflectivity and wave impedance parameters from seismic data under the constraints of interval velocity interfaces and borehole petrophysical parameters with nonlinear wave equation inversion technique. Finally, the resulting wave impedance was calibrated with CUSI neural network analysis to yield the absolute wave impedance of the formation between boreholes.
出处
《石油物探》
EI
CSCD
2002年第3期279-284,共6页
Geophysical Prospecting For Petroleum
