【摘要】 湘南多金属成矿带成为新一轮深部找矿的重点区域，电磁法作为深部多金属找矿的主要方法之一发挥着重要作用.然而湘南地区为湖南工业相对发达地区，前期大地电磁勘探显示该区强电磁脉冲等近场干扰严重(如风力发电设施、高铁运行等).针对该问题，本文提出一种时频结合的大地电磁去噪方法.在时域中，采用改进的变分模态分解方法与小波阈值去噪相结合以达到对大部分强电磁脉冲干扰的有效剔除.为了避免模态总数的人为主观选择，改进的变分模态分解方法引入加权峰度指数以自动确定模态分解总数.然而以上时间域方法对长时间电磁近场干扰剔除效果不佳，在频率域中这些噪声进一步通过形态聚类方法进行分离.该方法成功应用于湘南大地电磁勘探中，用于对强电磁干扰测站的去噪研究.首先重点分析了湘南大地电磁勘探中某一强电磁干扰测站的去噪效果，从多个角度分析本文提出的时频结合去噪方法的有效性.然后将该方法应用到其他强干扰电磁测站的去噪研究中，证明了本文提出的时频结合去噪方法的实用性.更多还原

【Abstract】 The polymetallic metallogenic belt in southern Hunan Province has become one of the most important regions for deep ore exploration. Electromagnetic(EM) method plays an important role in deep polymetallic ore prospecting. However, the southern Hunan region is a relatively developed area in Hunan province, and the early Magnetotelluric(MT) exploration shows that the near field interference such as strong EM pulse is serious(such as wind power generation, high-speed rail operation, etc.). To solve this problem, a time-frequency domain MT denoising method is proposed in this paper. In the time domain, the improved variational mode decomposition method(VMD) is combined with wavelet threshold denoising methods to effectively eliminate most of the strong EM pulse interferences. In order to avoid the subjective selection of the total number of modes, the improved variational mode decomposition method introduces weighted kurtosis index to automatically determine the total number of decomposition modes. However, the above time-domain method is not effective in eliminating the long-term EM near-field interference. In the frequency domain, these noises are further separated by morphological clustering method. This method has been successfully applied in the denoising study for the station with strong EM interference. Firstly, the denoising of one MT station from southern Hunan is carried out, and the effectiveness of the time-frequency domain denoising method proposed in this paper is examined. Finally, the proposed method is applied to another two MT stations with strong EM interference, which proves the practicability of the time-frequency domain denoising method proposed in this paper.更多还原