基于径向道变换和非稳态自适应匹配的线性相干噪声衰减技术.docVIP

 基于径向道变换和非稳态自适应匹配的 线性相干噪声衰减技术# 刘财，刘洋，王典* 5 10 15 20 25 30 35 40 （吉林大学地球探测科学与技术学院，长春 130026） 摘要：径向道变换是一种简单、有效的线性相干噪声消除方法，主要根据变换域中有效信号 和线性相干噪声具有明显不同的谱特征进行信噪分离，但由于其存在插值误差及参数选取等 问题，处理效果仍然不能令人满意。本文通过结合非稳态自适应匹配技术来提高径向道变换 的去噪能力。整形正则化提高了不适定性最小平方反演问题的求解速度，预测误差滤波器约 束条件的应用提高了目标函数的最小化匹配。通过分析理论模型和实际数据的处理结果，并 对比高通滤波、f-k 滤波以及最小平方匹配相减方法，本文方法得到了合理的线性相干噪声 消除结果，并且能够保护有效信号，更加适用于解决实际地震数据中非稳态的线性相干噪声 衰减问题。 关键词：径向道变换；非稳态；自适应匹配；线性相干噪声 中图分类号：P631.4 Linear coherent noise attenuation using radial trace transform and nonstationary adaptive matching LIU Cai, LIU Yang, WANG Dian (College of Geo-exploration Science and Technolgoy, Jilin University, ChangChun 130026) Abstract: Radial trace transform (RTT) is a simple method, which can effectively eliminate linear coherent noise. It can separate useful information and linear coherent noise according to more different spectra characteristics in RT transform domain. But the denoising result is not satisfied because of interpolation error and difficulty for choosing parameters. In this paper, a nonstationary adaptive matching technique that comes from multiple attenuation is introduced, which can effectively improve the denoising ability of RT transform. A kind of shaping regularization method gives much simpler and more effective approach for iterative problem. Prediction-error filter (PEF) provides the possibility for solving adaptive matching problem by using least-square method. Using synthetic and field-data examples and comparing our method with standard high-pass filtering, f-k filtering, and least-square matching subtraction, we demonstrate that the new technique can provide a better ability for coherent noise attenuation and signal protection. It is better fit for handling nonstationary seismic field data. Keywords: Radial trace transform; nonstationary; adaptive matching; linear coherent noise 0 引言 在反射法地震勘探中，地震线性相干噪声（包括面波、折射波和直达波等）属于规则干 扰，与随机噪声不同，其具有明显的运动学特征。线性相干噪声能量一般比较强，其消除结 ，2] 展了有效的线性相干噪声衰减方法，主要可以分为三类：第一类是基于数学变换的去噪方法， 如，经典的拉冬（Radon）