论文部分内容阅读
为了解决高斯分布约束地震反演结果在地层分界面附近存在振荡,且对地震数据中的噪声敏感,不利于识别隐蔽油气藏及小尺度地质体的问题,本文基于边缘保持平滑(EPS)算法,将非线性EPS算子作为正则化约束引入地震反演目标函数中,使反演的波阻抗剖面地层界面与地质体边缘更清晰,并且在对含有随机噪声的地震数据进行反演时,同样能够保持反演过程的稳健性.研究结果表明:非线性EPS正则化反演结果能更清晰地反映地层的变化,突出3~5m的小尺度岩性异常体,且在20%噪音背景下仍能够得到稳定的反演结果,具有较强的抗噪性,克服了高斯分布约束反演结果存在的振荡现象.实际地震资料的应用效果也表明,与高斯分布约束反演结果相比,边缘保持平滑正则化约束反演结果能够更好地反映地层的层状变化特点.
In order to solve the problem that the seismic inversion with Gaussian distribution constraints oscillates near the interface of stratigraphy and is sensitive to the noise in the seismic data, it is not conducive to the identification of subtle reservoirs and small-scale geological bodies. Based on the edge-preserving smoothing (EPS) algorithm, The non-linear EPS operator is introduced into the seismic inversion objective function as a regularization constraint, so that the interface of the impedance impedance profile stratigraphy and the edge of the geological body are clearer, and when inverting the seismic data containing random noise, And the robustness of the inversion process is maintained.The results show that the non-linear EPS regularization inversion can reflect the changes of the strata more clearly, highlighting 3 ~ 5m small-scale lithologic anomalous body and still capable of under 20% noise background The stable inversion result is obtained, which has strong anti-noise performance and overcomes the oscillation phenomenon existing in the inversion results of Gaussian distribution constraint.The application of actual seismic data also shows that the edge remains smooth compared with the result of Gaussian distribution constraint inversion Regularized constrained inversion results can better reflect stratiform stratigraphic change characteristics.