基于多项式拟合估算滑坡运动信息的研究  

Study on Estimation of Landslide Motion Information Based on Polynomial Fitting

作　　者：龚新亚[1] Gong Xin-ya

机构地区：[1]无锡南洋职业技术学院建筑工程与艺术设计学院,江苏无锡214000

出　　处：《建筑技术开发》2020年第19期155-157,共3页Building Technology Development

摘　　要：依据实测滑坡变形数据,通过中值滤波法对原始数据进行去噪处理,并对其处理后的数据运用基于偏差平方最小的最小二乘法进行多项式拟合,结果表明:此法的多项式拟合相关度较高,故可进一步对滑坡的变形轨迹进行预测,并估算滑坡的速度和加速度的大小及方向,进而得到目标较多的运动态势信息;通过计算机仿真也表明了当满足一定信噪比要求时,本方法具有较好的精度及稳健性,可为观测人员提供更多的数据分析,从而进行滑坡预测预警。According to the measured landslide deformation data,the original data is denoised by the median filter method,and the processed data are polynomial fitted by the least squares method based on the least square deviation. The results show that the polynomial fitting correlation of this method is high,so it can further predict the deformation track of the landslide,and estimate the velocity and acceleration of the landslide. The computer simulation also shows that the method has better accuracy and robustness when it meets the requirements of a certain signal-to-noise ratio,which can provide more data analysis dimensions for the observers,so as to carry out landslide prediction and early warning.

关 键 词：滑坡预警 变形 中值滤波 最小二乘法 多项式拟合 

分 类 号：TU42[建筑科学—岩土工程] P642.22[建筑科学—土工工程]