基于位移时间序列Fourier分析的滑坡预警研究

Research of landslide pre- warning based on Fourier Analysis for displacement- time sequence

滑坡灾害的发生是多种致灾因子共同作用的结果,滑坡位移-时间变化曲线蕴含了造成斜坡滑动位移的各影响因素的内在的趋势性及周期性变化的特征。将Fourier分析理论引入到位移-时间序列分析中,从频率域上来分析位移-时间序列在时间域上突变部分的频域特征,提出了滑坡预警频率域阈值的概念,建立了基于位移-时间序列Fourier分析的滑坡预警分级方法。该研究从频域的角度改进滑坡预警预报方法,从而达到更准确的预测预警效果。

The occurrence of landslide disaster is a result of co - action of multiple disaster - inducing factors. The landslide displacement - time sequence contains the internal characteristics of trend and periodicity of various inducing factors. The Fourier Analysis Theory is introduced into the analysis of displacement - time series. The frequency domain characteristic of displacement - time series mutation that is described in time domain is studied. The concept of landslide early - warning threshold on frequency domain is put forward, therefore a landslide early - warning classification methodology is built based on the Fourier analysis of displacement - time series. The research has improved the early warning and forecast method for landslide from the frequency domain.