【作者】 邱颖；

【导师】 席继楼；

【作者基本信息】 中国地震局地震预测研究所 ， 固体地球物理， 2008， 硕士

【摘要】 地电场观测对于探测地球内部电性、预测地震具有重要的现实意义。在电磁环境干扰日益严重的今天,我国地震电磁观测台站产出的地电场数据往往含有各种形式的干扰信号。排除这些干扰,成为地电场观测中必不可少的技术环节。本论文立足于软件滤波方法,基于地电场观测分钟采样数据以及通过试验采集的秒采样数据,研究了干扰形态规律比较清晰的电磁干扰信号的抑制和滤除方法,实现较强电磁干扰背景下的地电场信息的(准)实时识别和提取。本论文首先介绍了地电场信号的观测原理、地电阻率仪器供电干扰和地铁漏电干扰的产生机理和干扰特点等,对地电场信号及各种干扰信号的时域与频谱特征做了详细的分析。研究表明地电场信号与被研究的各种干扰信号在频率域上多有混叠,用常规的滤波方法不能滤除这些干扰。分析讨论了二进小波变换、多分辨分析、Mallat算法和小波域阈值滤波原理等。将小波域阈值滤波方法应用到地电阻率仪器供电干扰和地铁漏电干扰的抑制处理中,根据地电场信号与几种干扰信号的频谱特征确定了小波分解的层数和小波基。通过数值模拟实验,针对不同的干扰信号确定了不同的阈值选取规则,成功的抑制了地电场信号中的干扰成份。研究了基于LMS算法的自适应滤波方法和基于正交小波变换的小波域LMS自适应滤波方法。将其应用到地电阻率仪器供电干扰的抑制中,进行了数值模拟实验。由于常规的LMS算法具有结构简单,计算量小等优点,可以应用于滤除地电阻率供电干扰,通过大量实验确定了自适应滤波器的最佳参数。滤波结果表明,地电阻率仪器供电干扰信号得到了有效的滤除。而小波域LMS自适应算法得到的滤波输出信号与真实信号之间的的误差,明显小于利用常规自适应LMS算法所得的误差,收敛速度也有一定程度的加快。论文最后对小波域阈值滤波、自适应滤波等方法在实测地电场干扰处理中的应用进行了研究,主要包括高邮台地电阻率仪器供电干扰、静海台地铁漏电干扰、密山台类脉冲干扰等,实现了对实测受干扰地电场信号的降噪处理,达到了预期目标。并对其主要应用前景和当前存在的问题进行了分析讨论。本论文所讨论的分析和处理方法可以引申到地电场观测中的其他非工频干扰信号的分析和处理研究中,同时有望在地电场观测信息的提取和常规分析处理中获得应用。更多还原

【Abstract】 Geoelectric field observation has very important practical sense in detecting earth interior electrical structures and forecasting earthquake. The problems of electromagnetic environment interference are more and more serious in China. Geoelectric field .signal output by earthquake electromagnetism station always have varieties of noise jamming. Restraining these noises jamming becomes absolutely necessarily in geoelectric field observation. This paper is established in software filter method, based on minute sampling and second sampling geoelectric field data, research on restraining and filtering some regular interferences, implement real-time identification and extraction for geoelectric field signal from strong electromagnetic environment interference.This paper first introduced principles of geoelectric field observation and generation mechanism and characteristic of resistivity equipment artificial electricity interference, and metros leakage interference were especially investigated. The time domain and frequency domain character of geoelectric field signals and some interference signals were detailed analysed. Research shows that the investigated interference signals and geoelectric field signal are aliasing in frequency domain. These interferences can’t be filtered by conventional time domain or frequency domain filtering methods.Discusses the wavelet transformation elementary theory, and introduces the dyadic wavelet transform, multi-resolution analysis, Mallat algorithm and the theory of wavelet domain threshold filtering. Wavelet domain threshold filtering method is applied to restraining resistivity equipment artificial electricity interference and metros leakage interference. Based on characteristic of geoelectric field signals and the investigated interference signals,decompose layer and wavelet function are confirmed. In the wavelet domain threshold filtering, the selected threshold value direct impact on the filtering effect, and the traditional wavelet threshold selection rules do not apply to geoelectric field signals. By numerical simulation experiments, for different signal interference identified a different threshold selection rules, the interferences in geoelectric field signals were restrained successful. The basic principles of classical LMS self-adapting filtering algorithm and wavelet domain LMS algorithm were expounded and applied to restraining resistivity equipment artificial electricity interference. Then made some numerical simulation experiments, and he best adaptive filter parameters are determined. Filtering results showes that resistivity equipment artificial electricity interference is removed cleanly and effectively. It is also testified that wavelet domain LMS algorithm is more effective than classical LMS algorithm.Paper finally realize eliminating noise with the actual measurement geoelectric field observation signal, including resistivity equipment artificial electricity interference, metros leakage interference, pulse-like interference of Minshan station. Paper achieved the expected goals in practical disposal. And its main application and the existing problems were discussed in this paper can be extended to analysis and processing of other non-frequency interference signals in geoelectric field observations, as well as conventional analysis and information extraction of geoelectric field observations.更多还原