利用二进正交离散小波,通过Mallat快速算法对形变倾斜资料进行了处理,并将计算结果与该地区的地震活动进行了对比分析。综合分析认为,小波分解的本质是窗口加权滑动平均的结果,能有效地对各种不同频率成分的形变序列按不同的频段分离和重构,可以起到明显放大低频段趋势成份的作用,为有效识别形变趋势异常提供一种可行方法。
In this paper, tilt observation data of deformation has been calculated by using wavelet method of Mallat arithmetic, and the relation between gained result and seismic activity has been analyzed. The result indicates that the earth tilt data has obviously deformation abnormity before big earthquakes. Wavelet method has enlarged earth tilt abnormity in some frequency bands. So wavelet method gives us a strong tool in analysis of seismic precursor trend abnormity.