摘要
叠后地震反演方法很多,不同的反演方法具有不同的优缺点和用途。在叠后反演过程中,通常只选用一种方法进行一次反演,反演效果往往难以满足储层预测的精度要求。通过采用递推反演、约束稀疏脉冲反演和神经网络反演3种方法联合反演,不断提高反演结果的分辨率,增加信息量。反演获得高精度的波阻抗、储层地球物理特征参数等多种数据体,可为储层预测提供高精度的资料。实际应用表明,该反演技术流程是可行、有效和实用的。
There were many seismic pot-stack inversion methods with different characteristics and applied ranges.A single method was commonly applied to predict reservoirs,so it was hard to get a good result.Here a novel inversion workflow,which was different with traditional strategy,was proposed and discussed.Three post-stack inversion methods including recursive inversion,constrained sparse spike inversion and neural network inversion were jointed to increase the resolution,reliability of the result,and gain much information on reservoir rocks and fluid parameters during reservoir forecasting.The technique is adopted to integrate seismic,geological and well-log data to get the P-impedance data cube and other geophysical property data cube.The result shows that the new approach is feasible,effective and useful.It provides high precision data for reservoir prediction.
出处
《石油天然气学报》
CAS
CSCD
北大核心
2009年第5期246-249,共4页
Journal of Oil and Gas Technology
关键词
叠后地震反演
递推反演
稀疏脉冲反演
神经网络反演
储层预测
post-stack seismic inversion
recursive inversion
sparse spike inversion
neural network inversion
reservoir prediction
