%0 Journal Article %T 前兆观测干扰信号频谱特征分析 %T Analysis of frequency spectrum characteristics of disturbed signal in earthquake precursory observation %A 李继业 %A 武晓军 %A 刘长生 %A 任建辉 %A 高研 %A 孙鹏宇 %A 孙强 %A 秦丽岩 %A 战玉明 %A Li Jiye %A Wu Xiaojun %A Liu Changsheng %A Ren Jianhui %A Gao Yan %A Sun Pengyu %A Sun Qiang %A Qin Liyan %A Zhan Yuming %J 中国地震 %@ 1001-4683 %V 31 %N 3 %D 2015 %P 574-583 %K 前兆观测;干扰提取;谱分析;频谱特征 %K Earthquake precursory observation;Extract interference;Frequency;spectrum analysis;Spectrum characteristics %X 地震前兆观测数据中干扰信号频繁出现,大大增加了地震前兆异常信息提取难度。本文选用基于傅里叶变换的功率谱、傅里叶谱以及基于希尔伯特-黄变换的边际谱方法,针对黑龙江地区数字化水位和竖直摆倾斜测项中的典型干扰信号进行分析,提取了水位高频干扰、周期干扰以及竖直摆倾斜的风扰、同震应变等常见干扰信号的频谱特征。通过对比3种频谱分析方法的差异性,发现功率谱法和傅里叶谱对提取干扰和噪声等平稳信号更加有效,而边际谱法对提取同震应变波等非平稳信号的效果更加显著。 %X It is difficult to extract its anomaly information because the disturbed data is very complex in the earthquake precursory observation. In this paper, it analyses frequency spectrum characteristics of disturbed data about the high frequency interference,periodic disturbance of water level and wind,co-seismic strain of perpendicular pendulum tiltmeter,FFT and HHT marginal spectrum in Heilongjiang Province. The dominating vibration frequency is at over 1.7×10-3Hz in a period of lower than 10min. The three kinds of method have obvious differences. The power spectrum and FFT are effective methods in extracting noise and disturbed data. HHT marginal spectrum is an effective method in extracting coseismic strain. This study will enrich the theoretical application of spectrum analysis in extracting disturbed data. %R %U http://zgdz.eq-j.cn/zgdz/home %1 JIS Version 3.0.0