基于约束参数反演的共聚焦点道集偏移速度建模方法

Common focus point gather migration velocity modeling based on constraint parameter inversion

偏移速度建模的精度是影响偏移成像效果的重要因素。基于等时原理和差异时移(DTS)分析,提出了共聚焦点(CFP)道集偏移速度建模方法。聚焦算子的计算采用适于任意矩形网格的有限差分走时计算法,用波场延拓法进行CFP道集的生成,利用约束参数迭代反演实现偏移速度的更新,模型的参数化主要依据实际情况而定。为适应横向变速,选用了具有纵、横向速度梯度的参数化模型。模型的试算结果表明:(1)该建模方法具有较好的收敛性,一般只需迭代3～4次;(2)偏移速度的相对误差在1%范围内,反射层深度的误差小于20m;(3)合理选取参数化速度函数是该方法成功应用的关键;(4)要保证双聚焦的精度,必须合理选用计算聚焦算子和生成CFP道集的方法;(5)对复杂地质体的偏移速度建模一般只需分析和控制主要反射层就可以满足精度要求。

The accuracy of migration velocity modeling affects the result of migration imaging considerably. According to the principle of equal travel time and differential time-shift (DTS) analysis, a method of migration velocity modeling in common focus point (CFP) gather was put forward. The focusing operator was computed by finite-difference travel time method feasible to any rectangular grid and the CFP gather was generated by wave field continuation. The updating of migration velocity was completed by parameter constraint iterative inversion, and parameterization of model was determined mainly by real geological structure. To fit to laterally variant velocity, the parameterization model with vertical and lateral velocity gradients was selected. The test results prove that the velocity modeling has a better convergence, only 3 to 4 times iteration needed, and is more reliable with the migration velocity relative error of 1% and the reflector depth error less than 20 meters. Selecting the velocity function of parameterization is the key to its application. To guarantee the accuracy of double focusing, the focusing operator computing and CFP gather generating methods need to be selected properly. For complex geological body, only to analyze and control the main reflectors is enough to meet the required precision.