地震信号高分辨率时频分析方法及应用研究

发布时间：2018-05-11 17:53

本文选题：时频分析 + 高分辨率　；参考：《成都理工大学》2017年硕士论文

【摘要】：时频分析方法作为分析时变非平稳信号的有力工具,成为现代信号处理研究的一个热点,本文从时频分析理论出发,介绍了时频分析的基本理论和传统的几种时频分析方法,主要有短时傅里叶变换、小波变换、魏格纳分布等。与传统的时频方法相比,一种自适应时频分析方法希尔伯特-黄变换在处理分析非平稳信号具有明显的优势。但是在其核心算法EMD存在着模态混叠问题,为了解决此问题,研究了几种改进算法,包括集合经验模态分解和互补集合经验分解。虽然对模态混叠问题得到了一定的抑制,但是在实际应用中或多或少还是存在模态混叠现象,并且这类算法将原本简单的信号复杂化了,分离出过多的IMF分量,这就导致了本来的有效信号被过分的分割,可能使信号局部发生了畸变,或者说降低了信号的信噪比。经验小波变换算法是一种新型的时频分析方法。该算法打破了传统时频分析算法在自适应方面的局限性,结合了经验模式分解和传统小波变换的优势,可以将复杂信号分解为更具有物理意义的模式。本文深入研究了经验小波变换算法的原理,对算法中存在的问题进行改进和优化,验证了经验小波变换算法作为一种新的时频分析方法在实际应用中的价值。具体而言,针对该算法在处理复杂频谱的信号时出现的频谱划分问题,利用数学形态学在图像处理方面的优势,采用了基于morpho变换的经验小波变换算法;由于地震信号是非平稳复杂的信号,对信号的自适应性分割能力和在频谱中找到“有意义”模态的能力提出了更高的要求。因此,研究了基于尺度空间直方图分割的Otsu法,并应用于经验小波中的信号频谱的分割,最后得到了自适应经验小波变换。将自适应经验小波变换应用于地质正演模型与实际地震数据处理中,结果表明是一种高分辨率时频分析方法。通过经验小波变换得到的瞬时属性更加精确有效,其分辨率与可信度也大大提高。
[Abstract]:As a powerful tool for analyzing time-varying non-stationary signals, time-frequency analysis has become a hot topic in modern signal processing. Based on the theory of time-frequency analysis, this paper introduces the basic theory of time-frequency analysis and several traditional time-frequency analysis methods. There are mainly short-time Fourier transform, wavelet transform, Wigner distribution and so on. Compared with the traditional time-frequency method, an adaptive time-frequency analysis method, Hilbert-Huang transform, has obvious advantages in processing and analyzing non-stationary signals. But there is a modal aliasing problem in its core algorithm EMD. In order to solve this problem, several improved algorithms are studied, including set empirical mode decomposition and complementary set empirical decomposition. Although the modal aliasing problem is restrained to some extent, the phenomenon of modal aliasing still exists more or less in practical application, and this kind of algorithm complicates the original simple signal and separates out too many IMF components. This results in the over-segmentation of the original effective signal, which may make the signal local distortion, or reduce the signal-to-noise ratio (SNR) of the signal. Empirical wavelet transform (EWT) is a new time-frequency analysis method. This algorithm breaks the limitation of the traditional time-frequency analysis algorithm in the aspect of adaptation and combines the advantages of empirical mode decomposition and traditional wavelet transform to decompose complex signals into more physically meaningful patterns. In this paper, the principle of empirical wavelet transform (EWT) algorithm is deeply studied, the problems in the algorithm are improved and optimized, and the value of EWT algorithm as a new time-frequency analysis method is verified. Specifically, aiming at the spectrum partition problem in processing the complex spectrum signal, the paper uses the advantage of mathematical morphology in image processing, and adopts the empirical wavelet transform algorithm based on morpho transform. Because the seismic signal is non-stationary and complex, the adaptive segmentation ability of the signal and the ability to find the "meaningful" mode in the spectrum are higher requirements. Therefore, the Otsu method based on scale space histogram segmentation is studied, and applied to signal spectrum segmentation in empirical wavelet. Finally, adaptive empirical wavelet transform is obtained. Adaptive empirical wavelet transform is applied to geological forward modeling and actual seismic data processing. The results show that it is a high resolution time-frequency analysis method. The instantaneous attribute obtained by empirical wavelet transform is more accurate and effective, and its resolution and credibility are greatly improved.
【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】：P631.4

【参考文献】