基于压缩感知重构算法的大地电磁强干扰分离

为压制大地电磁信号中的强人文干扰,提出一种基于压缩感知重构算法的大地电磁信号去噪方法.通过构建与常见典型强干扰相匹配而对有用信号不敏感的冗余字典原子,利用改进的正交匹配追踪算法,分离出大地电磁信号中的强干扰成分.为了验证所述方法的强干扰分离效果,首先通过在实测大地电磁信号中加入理想的强干扰信号进行了仿真分离实验,然后从大量实测数据中选取三种含有不同类型强干扰的时间域片段,用所述方法对实测数据中的强干扰进行分离,最后将所述方法应用于青海试验点以及庐枞矿集区某测点实测数据的综合处理.仿真实验结果表明,该方法在分离出强干扰的同时,能够较好地保留有用信号.实测数据处理结果表明,该方法能够有效压制强干扰,改善强干扰区大地电磁数据的质量.

Strong noise separation for magnetotelluric data based on a signal reconstruction algorithm of compressive sensing

To suppress strong noise in raw magnetotelluric (MT) data, we propose a new time-series denoising method based on a signal reconstruction algorithm of compressive sensing. A redundant dictionary which matches with strong noise but insensitive to useful MT signal is designed, then the strong noise is separated from raw MT data by using this dictionary and an improved orthogonal matching pursuit algorithm. In order to verify the effect of the proposed method, firstly, a strong noise separation simulation experiment is carried out by adding ideal strong noise into the measured MT signal. Secondly, three time domain data segments with different types of typical strong noise selected from a large number of measured data are used for strong noise separation test. Finally, the proposed method is applied to measured data collected in the Qaidam Basin and Lu-Zong ore-concentration area. Simulation results show that the proposed method can effectively separate the strong noise from raw data while keep the useful part. The results of the applications to measured data show that our proposed method is an effective method to suppress strong noise and therefore improve the quality of MT data collected in areas with strong interference.