卡門率波應用於水準點位變動分析之研究

Abstract

經由監測水準點之變動可得知該點位之高程變化，進而獲悉該點位之地層垂直變動情形。 地層變動又是影響工程安全之重要因素，所以，對於其變動情形之分析實屬必要。對於點 位變動分析之方法可分為靜態分析與動態分析。靜態分析是根據觀測資料，對過去及目前 之地層狀況進行分析；而動態分析則加入時間因素，視地層變化量為時間之函數，故應具 有預估地層變動之能力。本文將介紹一種動態分析之方法："卡門濾波法"來進行水準點位 垂直變動分析。文中將分別介紹卡門濾波理論之推導，點位變動模式之建立，各項參數、 起始值之給定方法，最後以實際水準點位之觀測資料來作為本方法之分析實例。根據本研 究結果發現，在水準點位變動具有一定變化趨勢時，配合上適當合理的點位變動預估模式 ，將可利用此法有效的進行點位變動情形之預估。 The information about the vertical movements of the ground layers can be obtai ned by monitoring the vertical movements of the benchmarks . The status of the ground layers' vertical movements is one of the most important factors for en gineering safety . So it is necessary to analyze the movements of the ground l ayers . The methods for analyzing the vertical movements of the ground layers can be categorized into two divisions : "static analysis" and "kinematic analy sis" . In the kinematic analysis , the current heights of the benchmarks are c onsidered as the function of the velocity , the acceleration , and time , so t hey are possible to be predicted by the previous measurements . A method appl ied to analyze the vertical movements , called Kalman Filtering , is presented in this article . The process of the Kalman Filtering Algorithm is derived . The parameters of the models used in the prediction of the movements are assig ned . Beside , a case study on the implementation of this method is for refere nce .The results indicate that the movements of the ground layers are possible to be predicted by using Kalman Filtering Algorithm integrating the reasonab le models .