频率域滤波在重力异常分离中的研究及应用

发布时间：2018-03-10 22:20

本文选题：频率域　切入点：匹配滤波　出处：《成都理工大学》2017年硕士论文　论文类型：学位论文

【摘要】：在快速发展的新世代中,物探仪器精度的提高和理论基础的进一步完善促使物探方法也在快速发展。而物探方法的选取,取决于研究对象与周围物体的物理性质差异,基于密度差异的重力勘探更是不可或缺的技术之一。重力勘探包括数据采集,处理与解释三部分。重力场是叠加场,我们观测到的重力异常,是观测点下方各种不同密度体异常的综合反映,因此如何把我们研究(探测)的某些地质体产生的(局部异常)从整个叠加异常中准确地分离出来就显得特别重要。想要使分离的异常能更准确的表达出地质体情况,在尽量减少测量误差、降低仪器误差的基础上,而后期的数据分析、处理更是起了一个承上启下的作用,为异常解释提供准确的信息,这也是决定反演结果好坏的前提。从20世纪30年代重力仪的研制成功至今近100年时间,随着各国专家学者的不断研究与学习,对空间域方法,包括平均场法,圆周法,切割法等的文章已经屡见不鲜,对频率域滤波方法在异常分离中研究及应用写的较少,并且未给出与非常成熟的空间域方法的效果对比。本文细致讲述了频率域滤波方法中的匹配滤波和维纳滤波理论,然后编写了相关计算程序,通过理论模型实验及分析,论证了频率域滤波(匹配滤波和维纳滤波)对于重力叠加异常的分离效果,并与空间域方法平均场法,圆周法,切割法进行了对比。得出了同一模型下频率域滤波(匹配滤波,维纳滤波)分离异常的效果与空间域方法(平均场法,圆周法,切割法)近似,并在垂向叠加模型下分离异常效果优于空间域方法,在时间方面更是短于空间域方法。频率域滤波在分离异常中的实际意义在于可以从不同的频谱图中得出相应模型(矿体)的近似埋深信息(维纳滤波提取的是重力异常如径向最大对数功率谱图,匹配滤波提取的是重力异常垂向一阶导数径向平均对数能谱图),具有一定的自适应性。最后将研究结果应用到四川某铜铁矿重力叠加异常的分离中,取得了较好的效果,验证了该方法具有一定的实用性。
[Abstract]:In the new era of rapid development, the improvement of the precision of geophysical instruments and the further improvement of theoretical basis promote the rapid development of geophysical prospecting methods, and the selection of geophysical methods depends on the difference of physical properties between the objects studied and the objects around them. Gravity exploration based on density difference is one of the indispensable techniques. Gravity exploration consists of three parts: data acquisition, processing and interpretation. Gravity field is a superposition field. It is a comprehensive reflection of the anomalies of various density bodies below the observation point. Therefore, it is very important to separate the (local anomaly) produced by some geological bodies we study (detect) from the whole superimposed anomaly accurately. In order to make the separated anomaly more accurately express the geological body situation, On the basis of reducing the measurement error and the instrument error as far as possible, the later data analysis and processing play a connecting role in providing accurate information for the abnormal interpretation. This is also the premise of determining the inversion results. Since the development of the gravimeter in 1930s and the recent 100 years since the development of the gravimeter, with the continuous research and study of experts and scholars from all over the world, the spatial domain methods, including the mean field method and the circular method, have been studied. The article of cutting method and so on has been common, the frequency domain filter method in the unusual separation research and the application write less, In this paper, the matched filter and Wiener filter theory in frequency domain filtering method are described in detail, and then the related calculation program is written, and the theoretical model experiment and analysis are carried out. The separation effect of frequency domain filtering (matched filter and Wiener filter) on gravity superposition anomaly is demonstrated, and compared with the spatial domain method, mean field method, circle method and cutting method, the frequency domain filtering (matched filter) under the same model is obtained. The effect of Wiener filter is similar to that of the spatial domain method (mean field method, circumference method, cutting method), and is superior to the spatial domain method under the vertical superposition model. The practical significance of frequency domain filtering in separating anomaly is that the approximate buried depth information of the corresponding model (orebody) can be obtained from different spectrum maps (Wiener filter is the weight). Force anomalies such as radial maximum logarithmic power spectrum, The radial mean logarithmic spectrum of vertical first derivative of gravity anomaly is extracted by matched filter, which has certain adaptability. Finally, the research results are applied to the separation of gravity superposition anomaly in a copper and iron mine in Sichuan province, and good results are obtained. The practicability of this method is verified.
【学位授予单位】：成都理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：P631.1

【参考文献】