LLE重构和SVD分解的地震信号降噪方法

发布时间：2018-05-09 19:36

本文选题：局部线性嵌入 + 奇异值分解　；参考：《计算机工程与应用》2016年15期

【摘要】：针对现有地震信号降噪方法处理地震剖面的弯曲同相轴效果不佳,提出联合局部线性嵌入(LLE)和奇异值分解(SVD)方法的地震信号降噪技术。利用LLE的重构思想,对地震数据采样点用其近邻进行重构,实现非线性模式的弯曲同相轴的线性化处理,并去除一定程度的随机噪声;根据地震资料有效信号具有良好相关性的特性,采用SVD分解对LLE重构后的地震数据进行有效信号和噪声分离,剔除不相干的噪声,最终实现地震数据的随机噪声压制。在正演模型和真实地震资料上进行了实验,实验结果表明:与传统SVD方法相比,提出的方法很好地消除了随机噪声,有效信号基本上无丢失。
[Abstract]:Aiming at the poor effect of existing seismic signal de-noising methods in dealing with the bending cophase axis of seismic profile, a new seismic signal de-noising technique combining local linear embedding (LLE) method and singular value decomposition (SVD) method is proposed. Based on the reconstruction idea of LLE, the sampling points of seismic data are reconstructed by their nearest neighbors to realize the linearization of the bending cophase axis of nonlinear mode, and to remove the random noise to a certain extent. According to the good correlation characteristic of the effective signal of seismic data, SVD decomposition is used to separate the effective signal and noise from the reconstructed seismic data of LLE, and the irrelevant noise is eliminated, and finally the random noise suppression of seismic data is realized. Experiments on forward modeling and real seismic data show that compared with the traditional SVD method, the proposed method can eliminate the random noise and the effective signal is basically no loss.
【作者单位】：廊坊师范学院数学与信息科学学院;黑龙江八一农垦大学信息技术学院;
【基金】：黑龙江省自然科学基金(No.F200804) 黑龙江省教育厅科技项目(No.11531056)
【分类号】：P631.4

【相似文献】