摘要
在变形监测中获取的观测数据可以看作是时间与空间上的一组变形信号,一般该信号都会呈现趋势性,隐藏其中的周期性不易被发现;利用FFT对其拟合后的残差和小波分解后的高频信号进行变换与分析,通过时间序列分析对其分析结果建模得到短期的预测、预报,效果比较理想。
The observed data obtained in deformation monitoring can be regarded as a set of deformation signals in time and space,Generally the signal will show a trend, hidden in which the periodicity is not easy to be found; Transform and analysis of the high fre-quency signal after wavelet decomposition and the residual error after fitting by FFT; Through time series analysis, the results of the a-nalysis of the model are short term forecasting, the result is ideal.
出处
《测绘与空间地理信息》
2017年第4期173-175,178,共4页
Geomatics & Spatial Information Technology
关键词
变形监测
FFT
小波分解
时间序列分析
deformation monitoring
FFT
wavelet decomposition
time series analysis
