期刊文献+

常Q衰减介质分数阶波动方程优化有限差分模拟 被引量:5

Modeling of wave equation with fractional derivative using optimal finite-difference method in constant-Q attenuation media
下载PDF
导出
摘要 本文基于Kjartansson常Q模型理论,推导了常Q衰减介质中黏声波和黏弹性波的速度-应力方程,并采用基于二项式窗函数的优化交错网格有限差分方法进行了数值模拟,同时引入不分裂的复频移卷积完全匹配层(CPML)吸收边界条件,以消除边界反射.使用基于自适应时间步长记忆方法的中心差分近似时间分数阶导数,与常用的短时记忆方法相比,提高了波动方程的离散化精度和计算效率.通过对比均匀模型下声波的数值解与解析解,验证了算法的精确性,并进一步分析了不同品质因子下地震波的频散及衰减特征.对BP盐丘模型的数值模拟结果可以较好地反映本文数值方法对复杂介质的适应性及频散压制效果. In this paper,we derive the viscoacoustic and viscoelastic velocitystress wave equations in constant-Qattenuation medium.The optimal staggered grid finite-difference method based on binomial windows are used to numerically solve the equations,with incorporating convolutional perfectly matched layer(CPML)boundary conditions to eliminate boundary reflections.We introduce the adaptive time step memory method to approximate the time fractional derivative,which improves the discretization accuracy and computational efficiency of the wave equation comparing with the short memory method.Furthermore,we evaluate the accuracy of the algorithm by comparing the numerical solution with the analytic solution of the acoustic wave for the homogeneous media,and further analyze the dispersion and attenuation characteristics of seismic wave under different quality factors.Finally,we consider the BP salt model to demonstrate the applicability of our numerical algorithm in heterogeneous medium with obvious effect in suppressing numerical dispersion.
作者 孙成禹 乔志浩 伍敦仕 滕腾 Sun Chengyu Oiao Zhihao Wu Dunshi Teng Teng(School of Geosciences , China University of Petroleum (East China), Shandong Qingdao 266580, China CNOOC EnerTech-Drilling & Production Co, Tianjin 300450, China)
出处 《地震学报》 CSCD 北大核心 2017年第3期343-355,共13页 Acta Seismologica Sinica
基金 国家自然科学基金(41374123 41504097)资助
关键词 常Q衰减 分数阶导数 优化有限差分 自适应记忆 CPML constant-Q attenuation fractional derivative optimal staggeredgrid adaptive memory CPML
  • 相关文献

参考文献2

二级参考文献28

共引文献32

同被引文献31

引证文献5

二级引证文献20

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部