【作者】 龙云；

【导师】 吴燕冈；

【作者基本信息】 吉林大学 ， 地球探测与信息技术， 2010， 硕士

【摘要】 地震信号的匹配技术越来越受到人们的广泛重视,不管是地震数据的拼接成像还是时移地震数据处理,不同时间的地震数据匹配已成波形一致性处理的为核心技术。本文从信号的匹配滤波原理出发,从时间域和频率域分析了地震子波和地震数据的主要特征,研究了最小平方、相关等匹配准则,推导地震信号的匹配算法。以楔形模型和模拟数据为例,对频率域匹配滤波和小波域匹配滤波进行了对比。以单道记录和楔形模型为例,进行频谱分析,分析了在不同震源情况下产生不同子波对匹配滤波的效果,得出小波域的匹配滤波有很好的效果。利用小波分频特性,可以在匹配滤波的同时去除随机噪声,这是其他匹配滤波很难达到的效果。更多还原

【Abstract】 With the advances in seismic exploration, use different sources to match the filter so that the amplitude and phase of the data line, or in different times and comparative analysis of seismic data. Paper to simulate different focal positions generated in the same seismic data filtering, it can be seismic data analysis and processing, can be a very good analysis of seismic attributes, such as the relationship between geological structure, a series of hydrocarbon key issues. In order to achieve the effectiveness of how the removal of different sources of this type, different surface conditions, different construction age of various objective factors such as construction parameters caused by differences in seismic data, has become the key to solve this problem.This article will develop a mature combination of wavelet transform and matched filtering, focus on the matched filter based on wavelet transform in seismic data in different data source and application of different times. Especially for the matching portion of the seismic data are discussed in detail and compared with the Wiener filtering, highlighting the wavelet domain filtering, greatly improved signal to noise ratio. Matched filtering in wavelet domain has been largely improved, not only to improve the utilization of the data, but also improves the removal of random noise in the data. Not only matched filtering to achieve the best results in the match, but also to the original data signal to noise ratio has improved a lot. With random noise on seismic convolution model for the effective simulation experiments show that the method is effective.Fourier transform is an important basis for seismic exploration, this wavelet-based maturity based on the theory of wavelet sub-band features. Matched filtering technology, the prospects are very broad. This according to the advantages of both the two together effectively, using their properties to solve some practical problems.In this paper, based on previous years through the effectiveness of the method on the matched filter method for effective comparison and interpretation, use LS algorithm to solve the wavelet matched filtering problem, this paper, seismic modeling data to verify the advantages of wavelet transform matched filtering to the Fourier transform matched filtering, so that theoretically shows the matched filter based on wavelet transform has better precision. In addition, this paper, the seismic data as a research point of random noise, some features of wavelet Fourier change is unattainable, especially in wavelet decomposition and reconstruction, that is, frequency characteristics. This characteristic of wavelet matched filtering of wavelet domain stochastic noise in the process of removing the same time reaching the matched filtering can also be a large degree of increased noise ratio. This is the time domain and frequency domain matched filter matched filtering are unable to effect. Finally, the use of wavelet matched filter with a random noise signal and noise to match the reference Road, obtained wavelet matched filter can effectively remove the random noise. And other fx deconvolution, denoising compared to the removal of noise is greatly improved.更多还原