跳到主要內容

臺灣博碩士論文加值系統

(3.237.38.244) 您好!臺灣時間:2021/07/24 17:16
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:柯俊宏
研究生(外文):Chun-Hung Ke
論文名稱:運用貝氏網路建構不合格品質偵測輔助系統
論文名稱(外文):Developing the Bayesian Network Aided System for Non-confirming Quality
指導教授:謝孟勳謝孟勳引用關係
學位類別:碩士
校院名稱:國立中興大學
系所名稱:土木工程學系所
學門:工程學門
學類:土木工程學類
論文種類:學術論文
論文出版年:2008
畢業學年度:96
語文別:中文
論文頁數:77
中文關鍵詞:工程品質缺失偵測Bayesian Network
外文關鍵詞:Construction qualityLack and detect examiningBayesian Network
相關次數:
  • 被引用被引用:4
  • 點閱點閱:209
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:1
施工品質在整個工程生命週期中極為重要,如何及早發現缺失並加以改善,不但有助於延長工程使用壽命、亦可減少施工錯誤與風險。然而,茫茫工地現場中,工程查驗的項目常高達數十項甚至數百項。若要求品質查驗人員逐一檢驗所有可能的缺失項目,是件耗時費力的工作。如何在有限的人力、資源下,快速偵測出品質缺失所在,乃是本研究的核心目的。
貝氏網路(Bayesian Network)已廣泛應用於醫學界,其目的是儘量減少身體檢驗項目之原則下,並於追求節省醫療資源的前提下,仍可找出病人罹患何種疾病。另外,貝氏網路亦常被應用在垃圾郵件偵測,唯獨土木工程界尚鮮少藉由貝氏網路來協助偵測工程品質缺失。
本研究收集329件工程發生品質缺失之實際資料,結合Bayesian理論,訓練建置Bayesian Network。並自行研發出「貝氏網路輔助不合格品質偵測系統」(Bayesian network Aided system for Non-confirming Quality 簡稱:BANQ)。BANQ會依據品質缺失發生機率由高而低,對品質檢驗人員提出檢驗項目之建議。每當檢驗人員完成一項檢查項目並回饋檢驗結果之後,BANQ立即根據該項新增事實,立刻重新計算整體缺失的發生新機率,並再度建議下一項檢查項目。目前BANQ可於網路、平板電腦中運作,經驗證BANQ確實可協助品管人員,使用較少的檢查項目,達到相同的品管目的。
The construction quality is an extremely important issue. To early discover the construction flaw can not only extend the project service life, but also reduce the construction mistakes and risks. However, in construction sites, the number of inspection items is often as high as dozens or even several hundreds. How to detect and examine the construction quality effectively is the key purpose of this research.
The Bayesian Network theories have been widely applied to the medical arena to reduce body examination items and to save the medical resources. Also, the Bayesian Network theories are applied to detect junk mails. However, it is still not commonly applied to detect the construction project quality flaws.
This research collects data of quality defects from 329 recent construction projects. By applying the Bayesian Network theories, the data are trained to establish a Bayesian Network. The software is developed and called as “Bayesian network Aided system for Non-confirming Quality (BANQ). The BANQ first suggests for the inspectors the order of quality items need to be checked according to their probability from higher to lower. Once the suggested item is checked and the result of confirming or non-confirming is input into the system, the BANQ automatically re-calculates all the probability and immediately provides a new order of check-lists which follow the probabilities from higher to lower. Ten sets of testing data have been used to investigate the efficiency of the BANQ system and the result shows that the BANQ can reduce the inspection efforts by 44% in average.
第一章 前言 1
1.1 研究背景與動機 1
1.2 研究目的 1
1.3 研究範圍與限制 2
1.4 論文架構與研究流程 2
第二章 文獻回顧 4
2.1 文獻探討 4
2.2 關聯法則-支持度(Support Level) 6
2.3 貝氏網路理論(Bayesian Network) 7
2.4 貝氏網路結構學習 8
2.5 GeNIe 2.0使用介紹 10
2.6小結 10
第三章 工程品質缺失檢驗貝氏網路建立 11
3.1 貝氏定理(Bayes theorem) 11
3.2 貝氏網路推論範例 12
3.3 貝氏網圖之關係鏈結 13
3.4 貝氏網路研究流程 15
3.5 資料來源介紹及整理 15
3.6 以矩陣方式計算機率: 24
3.7 建置訓練Bayesian Network 30
第四章 成果分析 31
4.1 驗證原則 31
4.2 驗證貝氏網路 31
4.2.1驗證前人研究 37
4.2.2隨機抽樣檢驗: 38
4.2.3 無設計工項 40
4.2.3.1 不預先處理無設計工項 41
4.2.3.2 預先處理無設計工項類別 43
4.3 重複訓練 45
4.4 特例 45
4.5 BANQ 46
第五章 結論與建議 51
5.1 結論 51
5.2 未來建議 51
參考文獻 52
中文文獻 52
英文文獻 52
附錄A GeNIe使用介紹 53
附錄B驗證分析之結果 57
中文文獻
1.黃文宏,「資料採礦之Q-miner 應用於工程施工品質缺失檢驗」 ,中興大學,碩士論文,2002
2.張僩鈞,「兩階層式垃圾郵件過濾機制之研究」,2006
3.鄭重山,「貝氏網路之一致性修正比較及其實證研究」,清華大學,碩士論文,2000
4.蘇木春、張孝德,「機械學習:類神經網路、模糊系統以及基因演算法則」,1999
5.蘇俊豪,「建置貝氏網路於工程品質缺失及其應用之研究」 ,中興大學,碩士論文,2007

英文文獻
1.Bart Baesens,Stijn Viaene,Dirl Van den Poel,Jan Vanthienen,Guido Dedene,2001,「Bayesian neural network learing for repeat purchase modelling in direct marketing」
2.Bayes, T. , “An Essay Toward Solving A Problem in the Doctrine Of Chances”,Phil. Trans. 3:370-418,Reproduced in two papers by Bayes, ed. W. E. Deming,Hafner,New York,1763
3.Brenda McCabe,Simaan M. AbouRizk,Member,ASCE,and Randy Goebel,1998,「BELIEF NETWORKS FOR CONSTRUCTION PERFORMANCE DIAGNOSTICS」
4.Cooper, G., & Herskovits, H. “A Bayesian Method for the Induction of
Probabilistic Networks from Data. ”Machine Learning, 9,pp.309-347,1992
5.Daud Nasir, Brenda McCabe and Loesie Hartono “Evaluating Risk in Construction–Schedule Model .ERIC–S.: Construction Schedule Risk Model” JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT ASCE / SEPTEMBER/OCTOBER 2003,pp. 518-527
6.Eugene Charniak,1991,Bayesian Networks without Tears
7.Kluwer Academic Publishers, Boston,1992,「A Bayesian Method for the Induction of Probabilistic Networks from Data」
8.Pearl, J. “Probabilistic Reasoning in Intelligent Systems: Networks of
Plausible Inference. “ San Mateo, Calif.: Morgan Kaufmann. ,1988
9.Poole, David L., Mackworth, Alan, Goebel, Randolph G. Computational Intelligence: A Logical Introduction, Oxford University Press, New York ,
10.Pereira, C. A. D. B., and Barlow, R. E., “Medical Diagnosis Using Influence Diagrams”, NETWORKS, vol.20, pp. 565-577, 1990
11.Simon Haykin, Neural Networks: A Comprehensive Foundation, Prentice Hall,1999
12.Sucheta Nadkarni,Prakash P.Shenoy,1999,「A Bayesian network approach to making inferences in causal maps」
13.Thomas E. Tischer, P.E. and John A. Kuprenas, P.E. “Bridge Falsework Productivity—Measurement and Influences” JOURNAL OF CONSTRUCTION ENGINEERING AND MANAGEMENT ASCE /MAY/JUNE 2003,pp. 243-250
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top