基于广义S 变换时频滤波的MT数据去噪

针对油气勘探中大地电磁(MT)数据易受各类干扰的污染，且信噪难以分离的问题，把基于广义S变换的时频滤波技术应用于MT数据处理中来，得到MT数据的S域时频分布，分析受噪MT数据在S域的时频分布特征，再在S变换时频域进行时频阈值去噪，并对滤波后的S域时频谱进行逆变换重构，分离得到去噪后的MT数据。给出了基于广义S域时频滤波的方法原理与应用步骤，对被污染的仿真和实测MT数据进行了时频阈值滤波，并与小波阈值去噪方法进行了比较研究。结果表明：基于广义S变换的时频滤波方法可有效抑制MT数据中的干扰，从噪声信号中分离出有效的大地电磁数据，且减少了人为参与，提高了MT勘测的数据质量。

Denoising of MT data by time-frequency filtering based on the generalized S transform

This work attempts to solve the problem that magnetotelluric (MT) data in hydrocarbon exploration are prone to contamination by various interferences and it is difficult to separate signal and noise. The time-frequency filter based on the generalized S transform is applied to MT data processing. Firstly, the time-frequency distribution of the MT data in the S domain is obtained, then the time-frequency threshold denoising is performed, and the denoised time-frequency spectrum of the data is reconstructed by inverse transform. In this paper, the principle and method of the time-frequency filtering based on generalized S domain are given. The simulated and measured MT data contaminated by interferences are processed with time-frequency filtering. And the proposed method is compared with the wavelet transform method. Results show that the time-frequency filtering method based on the generalized S transform can effectively suppress the interference and separate the MT data from the signal bearing noise. The proposed method improves the quality of MT data.