• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文学位 >Wave equation travel time tomography based on the adjoint method.
【24h】

Wave equation travel time tomography based on the adjoint method.

机译：基于伴随法的波动方程行进时间层析成像。

获取原文
获取原文并翻译 | 示例
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Wave equation travel time Tomography based on the adjoint method is an advanced and useful theory to simulate the wave propagation and inversion. More practices and applications need to be implemented widely. A real seismic survey project is been done as an example in this paper with the above theory implementation. Also the comparison results can be funded in this paper.;Wave equation travel time Tomography based on the adjoint method is also a complex, long running, and repeated process. A desire to free hand from the inversion process needs to be met. Windows Workflow Foundation in .net 4.0 provides the feasibility to implement the inversion process automatically. The visual studio 2010 provides friendly platform for users to simply drag the function activity to design all kinds of workflow which also allowed both forward and backward workflow design. The benefits of the automatic implement the wave equation travel time inversion workflow is not limited to save the operating time, but also improves the inversion efficiency by reducing the un-continued running gap between each step. Furthermore, the automatic process provides the feasibility to implement other inversion process and reduce the seismic imaging period.;Wave Equation travel time tomography based on the adjoint method simulates the wave propagation path exactly in three-dimensional media which provides more accurate velocity model and much higher resolution image than the ray tracing theory which tracks the wave propagation path in two-dimensional media. That is also a big reason why the wave equation tomography is more accurate than the conventional tomography. Wave equation travel time Tomography based on adjoint method is not a new theory which was developed in 1980s but it was been limited by the requirements of large computing source which is used to calculate three-dimensional wave equations and to construct three-dimensional Fréchet kernel. Until recent years, following by the barge computing technology development, applying wave equation travel time tomography to large scale of earthquake and exploration become implementable.;In the practice of wave equation travel time tomography based on the adjoint method, forward wave-field and adjoint wave-field are been calculated by using source time function instead of storing green function. For adjoint wave-field, the wave field source is been derived by using reversed time signal at receiver as simultaneous source which is based on reciprocity of the green function property. And also all the synthetic seismogram relative to one shot can be simulated simultaneously for each wave-field calculation. Three-dementional Fréchet derivative is constructed by using both forward and adjoint wave-fields. Without storing green function, large disk space could be saved. In the meanwhile the simulation of two wave-fields will double the memory required. Since Hessian is also not available in the calculation, inversion process needs to be solved iteratively.
机译：基于伴随法的波动方程行进时间层析成像是一种模拟波传播和反演的先进实用理论。需要广泛实施更多的实践和应用。以上述理论实现为例，完成了一个真实的地震勘测项目。比较结果也可为本文提供依据。基于伴随法的波动方程行进时间层析成像方法也是一个复杂，耗时长，重复性好的过程。需要摆脱反转过程的渴望。 .net 4.0中的Windows Workflow Foundation提供了自动实施反转过程的可行性。 Visual Studio 2010为用户提供了一个友好的平台，用户只需拖动功能活动即可设计各种工作流，这也允许向前和向后的工作流设计。自动执行波动方程行进时间反演工作流程的好处不仅限于节省操作时间，而且还通过减少每个步骤之间的未连续运行间隙来提高反演效率。此外，自动过程为实施其他反演过程和减少地震成像周期提供了可行性。基于伴随方法的波动方程行进时间层析成像可以精确地模拟三维介质中的波传播路径，从而提供更精确的速度模型和更多信息。比在二维介质中跟踪波传播路径的射线跟踪理论更高分辨率的图像。这也是波动方程层析成像比常规层析成像更精确的重要原因。基于伴随法的波动方程行进时间层析成像技术并不是1980年代发展起来的新理论，但受到用于计算三维波动方程和构造三维Fréchet核的大型计算源需求的限制。直到近年来，随着驳船计算技术的发展，将波动方程行进时间层析成像技术应用到大规模地震勘探中成为可能。在基于伴随法，正向波场和伴随波的波动方程行进层析成像实践中波场是通过使用源时间函数而不是存储绿色函数来计算的。对于伴随波场，基于绿色函数特性的互易性，通过使用接收器处的反向时间信号作为同时源来推导波场源。而且，对于每个波场计算，可以同时模拟与一次发射有关的所有合成地震图。三向Fréchet导数是通过使用前向波和伴随波场构造而成的。如果不存储绿色功能，则可以节省大量磁盘空间。同时，对两个波场的仿真将使所需的存储空间增加一倍。由于Hessian在计算中也不可用，因此需要迭代求解反演过程。

著录项

  • 作者

    Zhang, Xiaofeng.;

  • 作者单位

    University of Wyoming.;

  • 授予单位 University of Wyoming.;
  • 学科 Geophysics.;Computer Science.
  • 学位 M.S.
  • 年度 2012
  • 页码 77 p.
  • 总页数 77
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号