分频段排干扰方法的建立与应用

Development and Application of the Eliminating Interfrence Method by Frequency-Band Division

在处理连续观测数据过程中,有时信号呈现出周期变化,如固体潮观测值中的潮汐信号等.直接对信号进行处理,如拟合、回归、调和分析等,是最常用的方法.当低频噪声背景和高频噪声均较强时,信号频率域(可称之为有用频段)内的信号被淹没了.直接提取信号较为困难.在这种情况下,本文试图利用分频段排干扰的方法,先将低频噪声和高频噪声分别排除掉,以便于进一步分析处理.本文实例计算表明,通过这样处理的应变固体潮观测资料,突出地显示了固体潮曲线的变化特征,而处理前的资料所对应的观测曲线,其固体潮汐的变化特征被噪声淹没了.

It is discovered in processing successive data that some data vary periodicaly such as tidal signals in the earth tide obervation. The methods for directly processing data such as fitting, regression, hamonic analysis, et al are used often. The signals in frequency domain (useful frequency band) will be submerged on condition that back ground of low frequency noise and high frequency noise are strong . and it will be difficult to direct separate the signals. In this case . this paper attempts to first eliminate the low and high frequency noises by means of the e- liminating interfrence method by frequency-band division in order that they will be analyzed and processed fur- therly . The practical examples in this paper show that the processed observed data of strain tide distinguished the feature of the earth tide , whill the feature of the earth tide in the observed curves responding to the unprocessed data has been submerged by the noises.