基于小波变换的地震信号拟合与去噪研究

发布时间：2018-02-23 20:47

本文关键词： 地震信号去噪小波变换人工拟合阈值小波包　出处：《河南理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：在实际信号的采集过程中,因为受到数据采集仪器和采集环境的影响,获得的地震数据难免会存在噪声,伴随在有用信号中的噪声是影响地震信号处理的重要因素,这些噪声会对有效信号的分析和处理工作带来不利影响,因此在不对有效信号产生损耗的基础上,尽最大的可能来降低信号含有的噪声,甚至从强噪声信号中提取有用信号,提高地震信号的信噪比对地震勘探精度的提高具有重要的意义,也是研究热点。本文主要研究了小波分析去噪的基本理论及其在模拟地震信号中的应用。首先分析了地震动噪声的特点和相应的去噪方式,通过分析和对比得出了傅里叶变换和短时傅里叶变换对于非平稳信号处理有些力不从心,而小波分析的方法却能同时对信号进行时域和频域的分析,适应多种复杂地信号分析状况,对于地震动这样的非平稳信号来说是很有力的处理工具。小波包变换更是对小波变换的进一步发展,对小波包空间作了更精细的分割,使小波包更具有更好的时频分析能力,特别是对于震动信号这种包含大量中、高频信息的非平稳信号能够进行更好的时频局部化分析,因此小波包去噪效果会更加明显。接着详细论述了地震信号的人工拟合方式,提出了基于小波包理论的地震信号模拟,通过比较得到了良好的非平稳地震信号。然后根据小波包阈值去噪原理提出了对新阈值函数的应用,并结合了加权中值去噪的方法,将此方法应用在模拟得到的地震信号中,通过计算比较重构误差,定量地分析了各种去噪方法的优劣。文中还探讨了地震信号小波变换的小波基选取问题。将小波分析去噪方法在实际地震信号去噪中进行仿真,取得了较好的效果,最后对基于Lab VIEW的上位机进行了设计,并将文中的算法在其中实现。
[Abstract]:In the acquisition process of the actual signal, because of data acquisition instrument and acquisition environment, the acquired seismic data will inevitably exist with noise, noise in a useful signal is an important factor affecting the seismic signal processing, the noise will adversely affect the analysis and processing of signals, resulting in not effective signal loss on the basis of the maximum extent possible to reduce the noise signal containing, and even extract the useful signal from strong noise signal, has an important significance to improve the signal-to-noise ratio of seismic signal to improve the precision of seismic exploration, is also a research hotspot. This paper mainly studies the basic theory of wavelet denoising and its application in earthquake simulation in the signal. The paper analyses the characteristics of ground motion noise and the corresponding denoising methods, through analysis and comparison of the Fourier transform and short time Fourier transform Ye Bianhuan for non-stationary signal processing there too, and analysis the wavelet analysis method can simultaneously in time domain and frequency domain of the signal, a variety of complex signal analysis for ground motion adaptation, non-stationary signal that is very powerful processing tools. Wavelet packet is the further development of wavelet transform, the wavelet packet space for a more precise segmentation, the wavelet packet has better time-frequency analysis ability, especially for the vibration signal contains a large, better time-frequency analysis of non-stationary signal to high frequency information, so wavelet packet denoising effect will be more obvious. And then it discusses the way of artificial fitting the seismic signal, is proposed to simulate the seismic signal based on wavelet packet theory, by comparing the obtained good non-stationary seismic signals. Then based on Wavelet Packet Threshold Denoising The application of the principle put forward a new threshold function, combined with the method of weighted median denoising of seismic signal, the application of this method in the simulation results, by comparing the reconstruction error, the quantitative analysis of the pros and cons of various denoising methods. This paper also discusses the selection of seismic signal wavelet transform wavelet base the wavelet denoising method. The simulation on the actual seismic signal denoising, and achieved good results, at the end of the host computer Lab based on VIEW is designed, and the algorithm in the implementation.

【学位授予单位】：河南理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN911.7

【相似文献】