网格逐次剖分算法在三维地震射线追踪中的应用

Application of grid successive subdivision algorithm in 3-D seismic ray tracing

初至波走时计算与路径追踪是基于射线理论层析成像的关键,走时线性插值(LTI)算法具有较高的追踪精度和计算效率,在二维条件下应用广泛,但在三维情况下存在最小走时方程不能解析求解的问题。为此,在走时线性插值的基础上,提出了一种改进的网格逐次剖分方法。该方法基于原有的LTI网格界面剖分算法,在最小走时剖分点附近建立同样规格的剖分点阵列进行进一步的细分搜索,寻找最小走时,避免了增加剖分精度而进行的全局细化搜索,节约了计算时间。该方法解决了三维LTI最小走时求解问题,并且能够以较小的计算代价获得更高的计算精度。在给出算法原理与计算步骤的同时,通过模型及实例验证了新算法在计算精度与计算效率上的改进。

First arrival traveltime calculation and path tracing is the key of tomography based on ray theory.Linear traveltime interpolation(LTI) algorithm has high tracing accuracy and efficiency,which has been widely used in 2-D condition.However,for 3-D problem,minimum traveltime equation cannot obtain analytic solution.Therefore,on the basis of LTI,an improved grid successive subdivision method was proposed.The method is based on the original LTI network interface subdivision algorithm,which carry out further search by establishing same mode subdivision point array close to the minimum-traveltime subdivision point to obtain minimum-traveltime.The method avoids global search caused by the increase of subdivision accuracy and saves computation time.Meanwhile,the method solves 3-D LTI minimum-traveltime computation problem and can achieve higher calculation accuracy with less computation amount.Model testing and actual data application results proved the modification of computation accuracy and efficiency.