三参数小波变换自适应阈值压制地震数据高频随机噪声

Suppression of high frequency random noise in seismic data by self-adaptive threshold of three parameter wavelet transform

在有效压制高频随机噪声的基础上,最大限度保留有效信号一直是地震数据降噪处理中的难点.小波变换阈值降噪因其优秀的降噪效果和较高的计算效率受到地球物理学家们的青睐.在小波变换阈值降噪中,小波基、阈值函数以及阈值是影响降噪效果的3个关键因素.针对这3个关键因素开展相关研究:(1)采用更加适合地震信号分解的三参数小波作为母小波,可以更好进行信噪分离;(2)采用改进的阈值函数克服传统硬、软阈值函数的不足,基于改进阈值函数得到的降噪地震数据不仅波形光滑、连续,而且基本保持原始信号的特征,综合了硬阈值和软阈值函数的优点;(3)针对统一阈值"过扼杀"有效信号这一缺点,提出自适应阈值,自适应阈值不单以能量为标准进行有效信号和噪声的识别,还考虑有效信号小波系数的相关性,避免有效信号被当成噪声舍弃.最后通过模型试算和实际地震数据验证所提方法的有效性和实用性.

On the basis of suppressing high-frequency random noise effectively,it is always difficult to retain effective signal to the maximum extent in seismic data denoising.Wavelet transform threshold denoising is favored by geophysicist because of its excellent denoising effect and high computational efficiency.In wavelet transform threshold denoising,wavelet basis,threshold function and threshold are the three key factors that affect the denoising effect.This paper carries out relevant research on the three key factors:(1)Three parameter wavelet,which is more suitable for seismic signal decomposition,is used as the mother wavelet to separate signal from noise;(2)Apply the improved threshold function to overcome the shortcomings of traditional hard and soft threshold functions.The denoised seismic data based on the improved threshold function not only has smooth and continuous waveform,but also keeps the characteristics of the original signal,which combines the advantages of hard threshold function and soft threshold function;(3)Aiming at the disadvantage of"over strangling"the effective signal by the unified threshold,a self-adaptive threshold is proposed.The self-adaptive threshold not only uses energy as the standard to identify effective signal and noise,but also considers the correlation of the wavelet coefficients of the effective signal,which effectively avoids the weak effective signal being abandoned as noise.Finally,the effectiveness and practicability of the proposed method are verified by model test and actual seismic data.