共聚焦点层析成像方法

面对复杂的地下空间，传统的层析成像方法存在诸多问题，比如在迭代求解的过程中需要反复进行偏移，计算量较大；利用射线追踪正演，地质结构复杂时射线反射角度难以求取。为此，提出了一种基于共聚焦点（CFP）理论进行速度反演的方法。该方法首先通过拾取DTS面板上的时移，迭代求取逆时聚焦算子；再利用逆时聚焦算子与时间残差，对速度模型的聚焦位置进行扰动；最后利用代数迭代方法中的系数将旅行时残差换算成慢度变化量，分配到射线路径上的每一个网格上，得到扰动后的速度模型；交替迭代更新聚焦点位置和网格速度以获得最终模型。与常规的速度反演方法相比，该方法利用单向旅行时进行反演，不需考虑复杂结构的反射问题，同时不用在每次迭代中重复偏移，并且只需考虑反射界面的数个几何点，避免了大型矩阵的求逆，从而大大减少了计算量。模型试算结果表明，即使在初始速度模型误差较大的情况下，也能快速收敛，取得较为满意的结果。

Research on tomography method based on common focus point

There exist some disadvantages in the traditional tomographic determination of seismic velocity model. For example, repeated migration process is required during the iteration to update velocity model that results in expensive computation. Besides, it is difficult to calculate the reflected ray angles when tracing ray through complicated geological medium. To tackle these problems, this paper proposes a velocity inversion method based on the theory of common focus point (CFP) technology. Firstly, the method obtains the time-reversed focusing operator using the time difference picked from the differential time shift (DTS) panel. Then, the positions of focusing points are corrected by using the focusing operator and time residual. Finally the velocity at each grid point are updated through the coefficients obtained in the algebraic reconstruction technique. The final model is obtained by alternatively updating the focusing positions and velocity at each grid point. Compared with the conventional velocity inversion method, the CFP method uses one-way travel time to solve the inversion without considering the reflection of complex structure and does not need to repeat the migration process in each iteration. This method avoids the inversion of a large matrix and only considers the groups of geometric points on the reflection interface, which thus saves computational time. The experimental results show that even if the initial velocity model is poor, fast convergence can be achieved, with satisfactory results obtained.