卡尔曼滤波在全张量磁梯度数据处理中的应用

发布时间：2018-05-11 12:49

本文选题：卡尔曼滤波 + 全张量　；参考：《中国地质大学(北京)》2015年硕士论文

【摘要】：矿产资源作为国家经济的重要战略资源,探明其储量及分布是一项重要工作。而在地质矿产探测方面,地球物理方法是一项重要手段。磁法勘探作为地球物理方法的一种,在区域构造推断、磁性岩体划分、岩层填图等方面发挥了重要作用。随着当代信息技术、仪器以及计算机技术的发展,航空磁法勘探逐渐成为磁法勘探的主流方向。本文深入研究了航空全张量磁梯度测量系统的基本原理,在理论研究的基础上完成相关正演程序的编写,系统的分析了测量过程中噪声的来源,针对噪声特征优选不同的滤波方法进行压制实验。一维剖面数据处理主要采用中值滤波和卡尔曼滤波进行,二维网格数据主要采用中值滤波、高斯平滑滤波以及卡尔曼滤波进行,同时以卡尔曼滤波为主要研究目标,通过模型试验对其规律函数进行选取,进一步完善其在全张量磁梯度数据中的应用。理论模型测试表明,使用卡尔曼滤波能够使噪声得到有效的压制,获得信噪比更高的数据。在实际应用中,选择上海市横沙岛实验全张量磁梯度数据,基本实现了上述方法在实际资料处理中的应用。基于以上研究内容,本论文主要取得以下几项成果:(1)研究了全张量磁梯度数据正演,编程实现了基本的正演,为后续数据预处理工作奠定基础;(2)针对全张量磁梯度测量特点,研究了已有的中值滤波、高斯平滑滤波及卡尔曼滤波方法,编程实现了相应的滤波器,并通过模型数据验证滤波方法的适(3)研究卡尔曼滤波基本理论,基本掌握了其基本性质及应用,建立模型验证其参数选取与其他方法滤波的效果对比得到相关的研究成果;(4)基于一维及二维数据滤波方法,编程实现数据处理可视化;(5)对比不同滤波器针对全张量磁梯度数据处理效果,得出相关结论。
[Abstract]:Mineral resources, as an important strategic resource of national economy, is an important work to prove its reserves and distribution. In geological and mineral exploration, geophysical method is an important means. As a kind of geophysical method, magnetic exploration plays an important role in regional tectonic inference, magnetic rock mass classification and rock mapping. With the development of modern information technology, instruments and computer technology, aeromagnetic exploration has become the mainstream of magnetic exploration. In this paper, the basic principle of aeronautical full Zhang Liang magnetic gradient measurement system is deeply studied, and the related forward program is compiled on the basis of theoretical research. The source of noise in the measurement process is systematically analyzed. According to the noise characteristics, different filtering methods are selected to carry out suppression experiments. One-dimensional section data processing mainly adopts median filter and Kalman filter. Two-dimensional grid data mainly adopts median filter, Gao Si smoothing filter and Kalman filter. At the same time, Kalman filter is taken as the main research object. Through the model test, its regular function is selected, and its application in the full Zhang Liang magnetic gradient data is further improved. Theoretical model tests show that using Kalman filter can effectively suppress noise and obtain higher SNR data. In the practical application, the Zhang Liang magnetic gradient data of Shanghai Hengsha Island experiment are selected, and the application of the above method in the practical data processing is basically realized. Based on the above research content, this paper mainly obtained the following achievements: 1) studied the forward modeling of the full Zhang Liang magnetic gradient data, programmed the basic forward modeling, laid the foundation for the follow-up data preprocessing work, and aimed at the characteristics of the full Zhang Liang magnetic gradient measurement. The existing median filter, Gao Si smoothing filter and Kalman filter are studied, and the corresponding filter is programmed. The basic theory of Kalman filter is studied through the model data validation. The basic properties and applications are mastered, and the model is established to verify that the selection of its parameters is compared with the results of other filtering methods, and the relevant research results are obtained, which are based on one-dimensional and two-dimensional data filtering methods. The visualization of data processing by programming is compared with the effect of different filters for full Zhang Liang magnetic gradient data processing, and the relevant conclusions are obtained.
【学位授予单位】：中国地质大学(北京)
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.24

【参考文献】