论文部分内容阅读
转换波的Kirchhoff叠前时间偏移算法由于受到算子假频的影响,会造成偏移成像质量降低.针对算子假频干扰成像这一问题,必须消除反算子假频的影响.文章对纵波算子假频生产的原理进行分析,避免了转换波反假频复杂公式和转换波参数,从反算子假频的基本公式出发,结合转换波偏移成像的公式进行推导引申,得到了转换波的反算子假频的数学表达式.在算法中利用了奈奎斯特采样定理和基于Z变换的复数数字滤波器设计方法,构建了局部三角形滤波器进行反算子假频处理.在算法实现过程中,创新之处在于利用CUDA的高性能计算能力,设计实现了高性能并行三角滤波算法.通过构建了一个脉冲模型验证了反算子假频算法,算法能够较好的去除脉冲中的算子假频产生的噪音,最后通过实际转换波资料的测试,利用CUDA的高性能计算实现的反算子假频算法,能够快速有效的消除算子假频的影响.
Kirchhoff prestack time migration algorithm of converted wave can reduce the quality of migration imaging due to operator aliasing, so the influence of inverse sub-aliasing must be eliminated in the case of operator aliasing interference imaging. Based on the principle of inverse sub-aliasing and combining with the formula of converted-wave migration imaging, we derive the derivation of In the algorithm, the Nyquist sampling theorem and the complex digital filter design method based on Z transform are used in the algorithm, and the local triangular filter is constructed to process the inverse sub-aliasing. In the process of algorithm implementation, the innovation lies in using CUDA’s high-performance computing ability to design and implement high-performance parallel triangular filter algorithm.By constructing a pulse model to verify the inverse sub-aliasing algorithm, the algorithm can better remove the impulse In the operator pseudo-frequency noise, and finally through the actual conversion wave data test, the use of CUDA high-performance computing to achieve anti-sub-sub-aliasing algorithm that can quickly and effectively eliminate Operator influence aliasing.