運用資料探勘技術於臺灣鋼筋混凝土橋梁構件劣化因子之研究; Data Mining the Deterioration Factors of Reinforced Concrete (RC) Highway Bridges in Taiwan

NCUIR > Engineering College > Graduate Institute of construction engineering > Electronic Thesis & Dissertation > Item 987654321/3075

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/3075

Title:	運用資料探勘技術於臺灣鋼筋混凝土橋梁構件劣化因子之研究;Data Mining the Deterioration Factors of Reinforced Concrete (RC) Highway Bridges in Taiwan
Authors:	李豪剛;Hao-Kang Lee
Contributors:	營建管理研究所
Keywords:	資料探勘（DATA MINING）;約略集(ROUGH SET);橋梁劣化;因子;Deterioration factors;Data mining;Rough set.;Bridge management
Date:	2007-06-08
Issue Date:	2009-09-21 12:04:27 (UTC+8)
Publisher:	國立中央大學圖書館
Abstract:	臺灣歷經數十年公路建設，交通網絡越趨完整，且因多山多河之特殊地景，橋梁遂為路網中不可或缺之要樞，根據「臺灣地區橋梁管理系統」統計，目前橋梁總數已達26,390座（2007年3月），其中橋齡逾20年以上者約達46.2％，可知多數橋梁已逐漸邁向老劣化之階段，但在人力及經費有限的情況下，如何以有限資源進行完善、有效率之橋梁維護管理，一直是公路主管機關所關心的議題。臺灣地區橋梁狀況於營運階段受到許多因素影響，如環境、交通、超載、設計及施工因素等，造成橋梁構件之劣化，進而影響橋梁發揮應有之功能，造成用路人安全性之疑慮。本研究根據既有國道橋梁構件之歷史檢測紀錄進行統計分析，由檢測紀錄中彙整橋梁構件損傷類型，如裂縫、混凝土剝落、白華等，並藉由資料探勘(DATA MINING)技術瞭解橋梁構件劣化之影響因子。本研究以約略集(ROUGH SET)理論為主要之探勘技術，利用分析軟體RSES(ROUGH SET EXPLORATION SYSTEM)透過法則推演技術以擷取大量資料庫中不精確、不確定與模糊之法則資訊，並從該法則資訊中擷取影響橋梁構件劣化之核心屬性，獲得影響橋梁構件各種劣化形式之影響因子，更藉由專家訪談驗證本模式中明顯與隱含知識之合理性。希冀經由本研究於橋梁構件劣化因子模式之建立，可將橋梁以所在地域、環境(如交通量、結構形式、距海距離等)之異同，進行更為精確之篩選分類，以挖掘類似工程及環境條件之橋群，繼而瞭解該橋群之劣化趨勢，使劣化預測模式得以透過歷史檢測記錄更符合實際現況，更可提供橋梁新建、養護機關於生命週期成本評估時較為客觀且準確之分析成果，進而決定較為適當之設計、維護管理與合理的預算編列，使橋梁使用年限得以延壽、預算養護效益達最大化之養護目標。 There are more than twenty-two thousands of highway bridges in Taiwan, with majority of them in the structural type of Reinforced Concrete (RC). As the average age of national bridge inventory grows older, more and more national funding are invested to maintain the entire bridge stocks in safe and sound operation. To pinpointing the deterioration factors of bridges will certainly assist in enhancing the effectiveness of bridge management. Conventionally the experience of senior bridge engineers is relied heavily upon for determining the deterioration factors of bridges. The literature shows a lack of a systematic way to determine the deterioration factors of bridges. Taiwan developed her own Bridge Management System in 1998. The basic bridge data as well as the inspection and maintenance data of more than twenty-two highway bridges are input in the system since. The objective of this study is to data mining the Taiwan Bridge Management System for the deterioration factors of RC highway bridges. However, instead of the twenty-one component types, only five component types, namely the expansion joint, deck, girder, abutment, and bearing, will be the subjects of the study for they have more maintenance records in the system. Rough Set theory, a data mining technique, and a RSES (Rough Set Exploration System) software system are employed in this study for data mining. As a result, major and minor deterioration factors of each of the five component types in RC highway bridges in Taiwan are identified. Meanwhile, statistical chi-square tests of each factor are conducted for the validation of the data mining results. Furthermore, the resulting deterioration factors are compared with those identified in the various previous studies in literature. The comparison results show that the deterioration factors identified in this study have a better interpretation of the deterioration in the BMS than the previous studies.
Appears in Collections:	[Graduate Institute of construction engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Size	Format

社群 sharing

Loading...