基于LMS算法的自适应滤波边坡位移预测  

Adaptive Filtering Prediction for Slope Displacement Based on LMS Algorithm

作　　者：王兴霞[1] 黄建文[1] 廖再毅[1,2] 

机构地区：[1]三峡大学水利与环境学院,湖北宜昌443002 [2]Department of Architectural Science,Ryerson University

出　　处：《人民黄河》2014年第9期101-104,共4页Yellow River

基　　金：国家重点基础研究发展计划项目(2012CB426502);三峡大学科学基金资助项目(KJ2012B027)

摘　　要：准确预测边坡位移可以有效判断边坡的安全状态和稳定性。针对边坡位移随时间的自适应变化特征,采用自适应滤波理论建立边坡预测模型,并利用最小均方(LMS)算法进行求解,同时给出了确定初始参数和模型检验的方法。在此基础上,将该模型应用于实际工程的边坡位移预测,研究结果显示,该模型的预测值与实测值吻合程度很高,具有良好的逼近能力和外推能力,以及较高的预测精度和可靠性。The slope’s safety state and stability can be effectively estimated by making an accurate prediction for slope displacement. Aiming at the characteristic of slope displacement’s adaptive variation with time,this paper applied the adaptive filtering theory to establish a kind of prediction model for slope displacement which could be solved by Least Mean Square (LMS)algorithm. At the same it provided the methods of initial parame-ters calculation and model testing. On basis of that,the model was applied to predict the displacement of a real slope by using the historical monito-ring data. The results indicate that the prediction data show a good agreement with the historical monitoring data,the prediction model gets a good approximation?capability and extrapolating?ability,and it has a higher prediction precision and a better reliability.

关 键 词：边坡 位移预测 自适应滤波 LMS算法 

分 类 号：TV541[水利工程—水利水电工程]