径向道域反褶积提高地震资料分辨率方法研究

发布时间：2018-05-17 14:56

本文选题：反褶积 + 径向道域　；参考：《吉林大学》2017年硕士论文

【摘要】：在我们平时生活的点点滴滴中我们都可以感受到,油气资源已经是我们生活中不可或缺的必须品。而对一个国家而言,油气资源的储备量和开发能力甚至可以影响其在国际上的竞争能力。因此提升地震勘探的能力已经是我国提升综合国力的一项重要任务。越来越深的勘探深度和越来越高的勘探精度需求使我们的地震勘探受到重重阻碍,给我们带来了极大的挑战。提高分辨率自然就是我们解决问题的关键。褶积又称为卷积,它和反褶积也可以叫去卷积是一种应用非常广泛的积分变换的数学方法。反褶积是我们处理地震数据时提高分辨率的一种常用处理技术,它是通过对地震子波的压缩,提取反射系数序列,把地震数据中的高频成份增强,同时将地震数据的频带给拓宽,最终达到提高分辨率的目的。但是典型的反褶积也存在很多的缺陷,比如它的假设条件多,算法不稳定等等。传统的非稳态褶积模型做出了地震波是垂直入射的假设,但实际在X-T域接收到的地震数据无法满足此条件,只有在0炮检距的情况下,地震波才能够满足自激自收的条件,并且当炮检距逐渐增大的同时,地震波越来越背离自激自收的条件。针对这个问题,本文将地震数据变换到径向道域中,使其在广义上满足垂直入射这个假设条件。本文首先介绍了分辨率的基础理论,和一些反褶积的方法,让大家对传统的方法有基本的认识,接着介绍了径向变换的原理,让大家能够理解将两种方法结合起的好处。在数据处理方面,本文首先构建了一个模型数据,接着对模型数据做了反褶积处理,得到处理后的结果。进而对原始模型数据做径向变换,在径向道域对其做反褶积处理,最后再将其反变换回X-T域。为了方便比较,我们分别做出了它们的频谱,重点将两个处理后的数据做对比。最终结果证明本文方法对模型数据的处理效果更佳。为了证明本文方法不只在理论上效果较好,我们又对一组实际数据做了相同的比较,最终证明本文方法对实际数据的处理效果依然很好。
[Abstract]:In our daily life, we can all feel that oil and gas resources have become an indispensable necessity in our lives. For a country, the reserve and development capacity of oil and gas resources can even affect its international competitiveness. Therefore, improving the ability of seismic exploration is an important task to enhance the overall national strength of our country. The deeper and deeper exploration depth and higher precision demand make our seismic exploration be hindered and bring great challenge to us. Improving resolution is, of course, the key to solving the problem. Convolution, also known as convolution, and deconvolution can also be called deconvolution is a very widely used mathematical method of integral transformation. Deconvolution is a common processing technique to improve the resolution when we deal with seismic data. It can enhance the high frequency components of seismic data by compressing seismic wavelet and extracting reflection coefficient sequence. At the same time, the frequency band of seismic data is widened to improve the resolution. But the typical deconvolution also has many defects, such as many hypothetical conditions, unstable algorithm and so on. The traditional unsteady convolution model assumes that the seismic wave is perpendicular to the incidence, but the seismic data received in X-T domain can not satisfy this condition. Only in the case of zero offset, the seismic wave can satisfy the self-excited self-collecting condition. At the same time, the seismic wave deviates from the self-excited self-receiving condition. In order to solve this problem, seismic data are transformed into radial channel domain to satisfy the assumption of vertical incidence in a broad sense. In this paper, we first introduce the basic theory of resolution and some deconvolution methods, so that we have a basic understanding of the traditional method, and then introduce the principle of radial transformation, so that we can understand the advantages of combining the two methods. In the aspect of data processing, we first construct a model data, then deconvolution the model data, and get the results after processing. Then the original model data is transformed in radial direction and deconvolution is done in radial channel domain. Finally, the original model data is inversely transformed back to X-T domain. In order to facilitate the comparison, we made their spectrum separately, and compared the two processed data. The final results show that the proposed method is more effective for model data processing. In order to prove that the method in this paper is not only effective in theory, we have also made the same comparison with a group of actual data. Finally, it is proved that the effect of the method in this paper is still very good.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P631.44

【相似文献】