金属矿地震资料降噪处理技术研究

发布时间：2018-05-26 06:08

本文选题：金属矿地震 + 信噪比　；参考：《中国地质科学院》2015年硕士论文

【摘要】：近地表矿产资源在快速发展的社会和经济的今天面临着日渐枯竭,这就使得深部勘探成为主流,金属矿的开采和勘探已经由浅层地表(500m)进入到第二开采和探测空间(500m-2000m,以及更深的探测空间)。由于金属矿地震的地质条件的复杂性,金属矿床形态(地质构造、控矿构造)十分复杂,地层倾角一般较陡；矿体的几何尺度一般较小；矿体与周围岩层之间的波阻抗差也很小；在矿体内部,呈现非均匀体性质,比如,矿体(如块状硫化物矿体)常常表现为次级散射源,其信号幅度弱,噪声常常将其淹没,通常使得在金属矿勘探中获得的地震记录信噪比较低。在金属矿地震资料处理中,地震资料处理工作中的一个最重要的环节是提高记录的信噪比和分辨率。噪声衰减是地震数据处理中非常重要的一项任务。目前的降噪处理方法对低速、低频干扰波的处理效果较好,对速度高、能量强的干扰波还缺乏有效的分离复杂地震波场和强有力的降噪处理手段。叠前降噪不干净严重影响了金属矿地震资料的处理效果。本文系统的介绍了金属矿地震数据的噪声来源,对震源噪声、外界噪声产生的原因、干扰类型和噪声特点等特性进行了详细介绍。在不同的数学域中,分析干扰波和有效波之间的差异,寻找差异最大的数学域,然后设计出合适的降噪算法进行信噪分离。如今去除地震波噪声已近有多种处理手段,Radon变换(线性Radon变换、抛物线Radon变换、双曲线Radon变换)为基础的降噪方法和技术,可使Radon域信噪更好分离,可以压制多次波和强噪声,分离转换波；基于小波分解和重建的干扰波降噪方法,实现噪声和信号的分离,这一变换适合面波和随机噪声的压制。本文设计了串联降噪技术及其处理流程,按照一定的顺序,对噪声进行衰减。着重介绍使用了小波变换和Radon变换及F-K滤波几种针对不同噪声的降噪处理方法。通过模型试算,进一步验证该降噪方法的可行性。金属矿地震数据中运用串联降噪方法,将会大大提高地震数据的信噪比,对金属矿地震数据中干扰多且能量强的噪声进行去除及压制都有较好的效果。在实际运用中,可以根据金属矿地震数据特殊性和差异性,选择不同串联降噪方法,以达到较好的处理效果(高信噪比和高分辨率的地震剖面),为后期工作提供保障。
[Abstract]:With the rapid development of society and economy, the near-surface mineral resources are increasingly depleted, which makes deep exploration into the mainstream. The mining and exploration of metal deposits has entered the second mining and exploration space from 500m to 2000m, and to the deeper exploration space from 500m to 500m. Because of the complexity of the geological conditions of metallic ore earthquakes, the shape of metal deposits (geological structure, ore-controlling structure) is very complicated, the stratigraphic dip is generally steeper, the geometric scale of orebody is generally small. The difference in wave impedance between the ore body and the surrounding rock is also very small; within the ore body, there is an inhomogeneous body, for example, the orebody (such as the massive sulfide orebody) is often shown as a secondary scattering source, the signal amplitude is weak, and the noise often submerges it. Usually, the signal-to-noise ratio (SNR) of seismic records obtained in metal ore exploration is low. In the processing of metal ore seismic data, one of the most important steps in seismic data processing is to improve the signal-to-noise ratio and resolution of the records. Noise attenuation is a very important task in seismic data processing. The current noise reduction methods have good effect on low-speed and low-frequency interference waves, but lack of effective separation of complex seismic wave fields and powerful means of noise reduction for high velocity and strong energy interference waves. The unclean pre-stack noise reduction seriously affects the processing effect of metal ore seismic data. In this paper, the noise sources of metal mine seismic data are systematically introduced. The source noise, the causes of external noise, the types of interference and the characteristics of noise are introduced in detail. In different mathematical domains, the difference between the interference wave and the effective wave is analyzed, and the mathematical domain with the biggest difference is found, and then an appropriate denoising algorithm is designed to separate the signal and noise. Nowadays, there are many methods and techniques to remove seismic wave noise based on Radon transform (linear Radon transform, parabola Radon transform, hyperbolic Radon transform), which can separate the signal noise in Radon domain better and suppress multiple waves and strong noise. The noise is separated from the signal by the noise reduction method based on wavelet decomposition and reconstruction, which is suitable for the suppression of surface wave and random noise. In this paper, the series noise reduction technology and its processing flow are designed, and the noise is attenuated in a certain order. Wavelet transform, Radon transform and F-K filter are used to reduce noise. The feasibility of the noise reduction method is further verified by the model test. The use of series noise reduction method in metal mine seismic data will greatly improve the signal-to-noise ratio of the seismic data, and it has a good effect on removing and suppressing the noise in the metal mine seismic data which has a lot of interference and strong energy. In practical application, different series noise reduction methods can be selected according to the particularity and difference of metal mine seismic data, in order to achieve better processing effect (high signal-to-noise ratio and high resolution seismic profile), and provide guarantee for later work.
【学位授予单位】：中国地质科学院
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P631.44

【参考文献】