利用压缩感知方法提高地震资料分辨率

发布时间：2018-04-14 03:01

本文选题：压缩感知 + 稀疏化　；参考：《石油地球物理勘探》2017年02期

【摘要】：将压缩感知理论引入地震资料处理,研究了基于压缩感知理论的提高地震资料分辨率方法,针对压缩感知理论中的信号稀疏化、观测矩阵设计、信号重构等流程提出了改进措施。针对传统观测矩阵取高斯随机矩阵后计算结果不稳定问题,根据地震记录的信号特点,将高斯随机矩阵改进为较大反射信号约束下的观测矩阵,提高了重建算法的速度和精度;为了压制局部均匀化和区域均匀化噪声,增强较弱的有效信号,研究了地震资料的规格化处理方法,提出了局部压缩感知和区域压缩感知联合处理的方法,增大了源信号独立性条件范围。模型数据实验和实际数据应用效果表明,基于压缩感知理论的提高地震资料分辨率方法使弱信号增强、地震分辨率明显提高。
[Abstract]:The compressed sensing theory into seismic data processing, research to improve the resolution of seismic data method based on compressed sensing theory, the compressed sensing theory of sparse signal, design of measurement matrix, signal reconstruction process improvement measures are put forward. According to the traditional observation matrix from Gauss random matrix after the calculation results is not stable, according to the signal characteristics the seismic records, the Gauss random matrix is improved greatly under the constraint of the observation matrix reflection signal, improves the speed and accuracy of the reconstruction algorithm; in order to suppress the uniform local and regional uniform noise, enhance the effective weak signal, the seismic data normalization processing method, we propose a local compressed sensing and regional compression method the combined treatment increases the perception, source signal independence conditions. Experimental results show that the model data and actual data compression based on application effect. The enhancement of seismic data resolution by perceptual theory makes the weak signal enhanced and the resolution of the earthquake increases obviously.

【作者单位】：中国石油大学(华东)地球科学与技术学院;
【分类号】：P631.44

【相似文献】