標題: 衰退資料的非線性模式
On Nonlinear Modeling for Degradation Data
作者: 劉碧惠
Liou, bie-huei
李昭勝, 劉淑鶯, 趙蓮菊, 彭南夫
Jack Lee
統計學研究所
關鍵字: 成長曲線模式;衰退資料;冪次轉換;非線性估計;隨機效應;AR(1);Growth Curve Model;Degradation Data;AR(1);Box-Cox power transformation;Nonlinear esitmation;Random effect
公開日期: 1997
摘要: 這篇論文用非線性的成長曲線模式來分析 Bogdanoff 和 Kozin (1985)兩人所提供的裂縫長度資料. 這篇論文主要的目的是用預測觀點來 比較非線性的成長曲線模式與 Lu 和 Meeker (1993) 所提的非線性衰退 模式及 Lee andLee (1998) 的線性成長曲線模式. 另外, 我們也用這些 不同的模式來做斷裂時間點的預測. We use the nonlinear growth curve model to analyze the crack lengthdata from Bogdanoff and Kozin (1985) in this thesis. The main purpose ofthis thesis is to compare the proposed models with the nonlinear degradationmodel of Lu and Meeker (1993) and the linearized growth curve model of Leeand Lee (1998) from a prediction point of view. Besides, we also use differentmodels to compare the forecasts of the failure time.
URI: http://140.113.39.130/cdrfb3/record/nctu/#NT860338012
http://hdl.handle.net/11536/62707
顯示於類別:畢業論文