基于组合滤波的矿集区大地电磁信号去噪

发布时间：2018-02-27 08:12

本文关键词： 经验模态分解数学形态学滤波自适应滤波大地电磁去噪矿集区　出处：《吉林大学学报(地球科学版)》2017年03期 　论文类型：期刊论文

【摘要】：针对矿集区大地电磁(MT)信号受环境噪声和人文噪声污染严重的问题,提出一种结合了经验模态分解(EMD)和数学形态学滤波的组合滤波方法,对矿集区大地电磁信号的时域信号进行滤波处理。介绍了方法原理和计算步骤,评估了该方法的去噪效果;在与小波变换去噪效果对比的基础上,用仿真实验验证了方法的可靠性,并对某矿集区的实测数据进行了去噪处理。结果表明,组合滤波方法充分利用了EMD多尺度分解及其可重构特性和数学形态学滤波方法的优点,在滤除噪声的同时为MT信号尽可能多地保留了有用信息。去噪后,估算的响应曲线方差减小到原来的一半,为进一步正确资料处理和地质解释提供了保障。
[Abstract]:Aiming at the serious environmental and human noise pollution of magnetotelluric magnetotelluric (MTT) signals in ore concentration areas, a combined filtering method combining empirical mode decomposition (EMD) and mathematical morphological filtering is proposed. In this paper, the time domain signal of magnetotelluric signal in ore concentration area is filtered. The principle and calculation steps of the method are introduced, and the denoising effect of this method is evaluated, which is compared with that of wavelet transform. The reliability of the method is verified by simulation experiments, and the denoising processing of the measured data in a mine concentration area is carried out. The results show that the combined filtering method makes full use of the advantages of the EMD multi-scale decomposition and its reconfigurable properties and the mathematical morphological filtering method. After denoising, the variance of the estimated response curve is reduced to half of the original one, which provides a guarantee for further correct data processing and geological interpretation.
【作者单位】：湖南文理学院物理与电子科学学院;中南大学地球科学与信息物理学院;
【基金】：国家自然科学基金项目(41304098) 湖南省自然科学基金项目(2017JJ2192) 湖南省教育厅重点项目(16A146)~~
【分类号】：P631.325

【相似文献】