基于Hilbert-Huang变换的大地电磁数据处理方法研究

项目名称： 基于Hilbert-Huang变换的大地电磁数据处理方法研究

项目编号： No.41304098

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 天文学、地球科学

项目作者： 蔡剑华

作者单位： 湖南文理学院

项目金额： 25万元

中文摘要： 寻求现代信号处理方法来处理大地电磁（MT）数据以获得"无偏"阻抗估计，是地球物理工作者研究MT测深的重要课题。我们的前期研究表明：用新发展的非线性信号处理方法Hilbert-Huang变换（HHT）来估算功率谱更切合MT信号非平稳特性的实际。这为减小电法勘探中由于MT信号的非平稳特性与传统 Fourier 变换方法不适宜带来的估算偏差提供了思路。本项目拟在前期研究基础上，采用与传统Fourier 变换对比的方法，系统研究由Hilbert时频谱统计估计MT信号功率谱，再由此功率谱来估算响应参数的大地电磁数据处理方法。重点研究功率谱估计及提高谱估计精度的方法；建立响应参数估算的数学模型，以及新方法的评价体系；编制一套可视化软件以适应实测MT数据处理的工程需要。最终建立一套系统的基于HHT的MT数据处理方法，为最小化传统方法带来的估算偏差提供技术支持，为后续精确的定量反演奠定基础。

中文关键词： 大地电磁数据；Hilbert-Huang 变换；评价体系；功率谱估计；响应函数估算

英文摘要： For MT sounding scholars, it is an important topic to seek modern signal processing methods to process the magnetotelluric (MT) data for obtaining the no estimation bias impedance.Our preliminary study proved that spectrum estimation method, which was based on the Hilbert-Huang transform(HHT), could fit the non-stationary characteristics of MT signal better. HHT provide a new road to overcome the deviation of estimation which is caused by the fact that the non-stationary characteristic of MT signal is not suitable for the Fourier transform method. In this project,based on previous research the Hilbert-Huang method will be used to process MT data and compared with Fourier transform. It will be deeply studyed how to use the Hilbert spectrum to statistically estimate power spectrum of MT signal and to solve for response function further.The evaluation system of new method will be builded. Some key technology will be focused on and discussed in detail. One is the power spectrum estimation and the method to improve the accuracy of estimation. The second is to establish the mathematical model and the evaluation system of response function estimation. Lastly, a MT data processing system based on Hilbert-Huang transform will be established and a set of visualization software will be developed. The research findings can

英文关键词： Magnetotelluric data；Hilbert-Huang transform；evaluation system；power spectra estimation；response function estimation