论文部分内容阅读
常规逆时偏移可以实现较好的构造成像,但由于照明不均等因素使得该方法不能实现对岩性储层的精确刻画。为了得到可靠的地下反射界面的反射系数,需要用反演的方式解决成像的问题。最小二乘逆时偏移(LSRTM)被称为线性反射率反演,它通过引入Hessian矩阵实现相对的高分辨率振幅保真成像。共轭梯度算法是非常高效的迭代算法,使得LSRTM方法变得实用。基于模型数据与观测数据的互相关程度判定速度模型的准确度及计算模型更新量,可以使得LSRTM摆脱地震子波的依赖,增强稳健性。从模型试算及实际资料处理中可以看出,相比常规RTM和单程波偏移方法,LSRTM的成像结果可以直接应用到后续的储层描述和四维地震中。本论文主要研究了最小二乘RTM的一阶近似,也就是线性Born近似。当遇到更复杂的地质构造时,可以通过考虑更高阶的近似来提高其应用效果。
Conventional inverse time migration can achieve better imaging of the structure, but due to the uneven illumination, this method can not accurately describe the lithologic reservoir. In order to obtain a reliable reflection coefficient of the subsurface reflection interface, the problem of imaging needs to be solved by means of inversion. Least Squares Inverse Time Shifts (LSRTM), known as linear reflectivity inversion, achieve relative high-resolution amplitude-fidelity imaging by introducing Hessian matrices. The conjugate gradient algorithm is a very efficient iterative algorithm that makes the LSRTM method practical. Based on the degree of cross-correlation between the model data and the observed data, the accuracy of the velocity model and the update of the model are calculated to make the LSRTM get rid of the seismic wavelet dependence and enhance the robustness. As can be seen from the model trial calculation and actual data processing, compared with the conventional RTM and one-way wave migration methods, the LSRTM imaging results can be directly applied to the subsequent reservoir description and 4-D seismic. This thesis mainly studies the first-order approximation of least square RTM, which is also the linear Born approximation. When encountering more complex geological formations, the application effect can be improved by considering higher order approximations.