基于形态滤波和HHT的地磁信号分析与预处理  

Analysis and preprocessing of geomagnetic signals based on morphological filter and Hilbert-Huang transform

作　　者：李季[1] 潘孟春[1] 唐莺[1] 罗飞路[1] 

机构地区：[1]国防科学技术大学机电工程与自动化学院,长沙410073

出　　处：《仪器仪表学报》2012年第10期2175-2180,共6页Chinese Journal of Scientific Instrument

基　　金：国家自然科学基金(51175507)资助项目

摘　　要：在地磁导航技术中,为获得精确的地磁场值,必须对信号中含有的各种噪声进行处理。针对地磁测量信号中含有的噪声特点,提出将形态滤波与Hilbert-Huang变换结合起来,先用形态滤波对数据进行处理,消除瞬时强脉冲的影响,再将信号进行经验模态分解,作出其二维时频谱图,分析信号中含有的噪声的特点,并根据分解的结果确定阈值,对各分量进行低通滤波,然后再重构。通过实际采集的数据对所提出的方法进行了验证,结果表明提出的方法可以有效地对地磁测量信号中的各种噪声进行处理。In geomagnetic navigation, in order to get accurate measurement values of geomagnetic field, various kinds of noises must be eliminated beforehand. Through analyzing the characteristics of the noises in the geomagnetic measurement signals, this paper presents a new method that combines morphological filter and Hilbert-Huang transform（HHT） . The data is processed firstly using morphological fil- ter to eliminate the influence of the instantaneous strong pulses, then the processed data is decomposed using empirical mode decomposi- tion （EMD） method, and the two-dimensional Hilbert frequency-time spectrum is achieved. The characteristics of the noises in the signal are analyzed, and a threshold is decided according to the decomposition results; the decomposed components are low-pass filtered on main intrinsic mode functions （ IMFs）, finally a reconstruction is carried out to get the preproeessed signal. An experiment on the actually acquired data was carried out to verify the presented method, and the result shows that this method can effectively process and eliminate the noises in geomagnetic measurement signals.

关 键 词：地磁导航 地磁信号 形态滤波 HILBERT-HUANG变换 预处理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术] TH76[自动化与计算机技术—计算机科学与技术]