曾惠斌臺灣大學：土木工程學研究所鍾金龍Chung, Chin-LungChin-LungChung2007-11-252018-07-092007-11-252018-07-092007http://ntur.lib.ntu.edu.tw//handle/246246/50464通常橋梁維護管理系統對橋樑之劣化及其功能績效之分析與評估，主要取決於目視安全檢測之完整性與正確性。然而目視安全檢測在執行過程中都相當主觀與不確定性，往往不適當之檢測資料造成後續分析橋梁狀況得到不當的結果與對策。本研究希望藉由大量之目視安全檢測資料以統計之處理手法做有系統之分類及分析，以達到下列幾項目標: (1) 分析橋梁及其各構件之劣化程度; (2) 根據績效指標決定橋梁及其各構件之壽命劣化趨勢; (3) 並依歷年經驗值研判各橋梁安全性構件一般預防性養護與重大維護之機制 為達到上述之目的，本研究從巨觀及微觀之角度切入，並以DER經驗公式及採資料採礦分析U值等兩種方式做比較，最後綜合提出最適合國內參考之劣化模式與維護機制。Usually the rating values are determined by visual inspection in most of the Bridge Management Systems (BMS). A visual inspection is not very objective, and can sometimes result in a different assessment of element condition what may lead to the improper prioritization for taking corrective action. Our novel evaluation results in a more objective condition rating based on the visual inspection data. In order to provide an easier way to arrange and prioritize the maintenance schedule among the bridges, the health condition of a bridge needs to be diagnosed in a more practical and more accurate manner. This work utilized systematic classification and statistical analysis based on the thousands of sets of existing bridge inspection data collected in Taiwan over the past 5 years, with the following goals in mind: (1) develop a normal zone in the health condition shown on the diagnostic figure, and identify the health situation for the bridge using a specific inspection condition; (2) determine the time when the bridge will be in the worst situation, then determine the proper time for maintenance to be carried out; (3) introduce a new data process and innovative model into the BMS in Taiwan, as it will definitely be an asset to be able to estimate reliable maintenance costs in the future. In factor, the status of a bridge is determined by its structural capability and material strength. Consequently a lot of researchers have studied the failure, the fatigue, and the deterioration of the structure in terms of the structural function of a bridge. However, the overall performance of a bridge may be affected simply by the damage of one of its components. Therefore this study utilized a systematic classification and statistical analysis based on the existing bridge inspection data collected in Taiwan to reach the following goals: (1) assess the performance distribution and deterioration rate for components of bridge; (2) find out the right time to do the preventive and essential maintenance for the component of bridge with an empirical method, and to decide what time and which component of a bridge will receive preventive maintenance or regular maintenance. Finally the application of data mining was introduced to respond the reliability of the results by comparison methods from macro and micro point of views. Key Words: Inspection, Performance, Bridge Maintenance, Deterioration, Safety, Data Mining.ACKNOWLEDGEMENT...........................................i ABSTRACT IN CHINESE......................................ii ABSTRACT IN ENGLISH.....................................iii TABLE OF CONTENTS.........................................v LIST OF FIGURES........................................viii LIST OF TABLES..........................................xii CHAPTER 1 INTRODUCTION...................................1 1.1 BACKGROUND......................................1 1.2 LITERATURE REVIEWS..............................2 CHAPTER 2 RESEARCH OBJECTIVES............................6 CHAPTER 3 METHODOLOGY...................................10 3.1 DATA TREATMENT.................................20 3.2 RANKING SYSTEM AND SCORING.....................25 CHAPTER 4 MODELING AND ANALYSIS.........................38 4.1 MACRO DETERIORATION MODEL......................38 4.2 MICRO DETERIORATION MODEL......................45 4.3 ANALYSIS.......................................50 CHAPTER 5 MAITENANCE SCHEDUL............................52 5.1 HEALTH DIAGNOSIS AND MACRO MAINTENANCE STRATEGY................................................52 5.2 MICRO MAINTENANCE SCHEDULE.....................54 5.3 SUMMARY........................................58 CHAPTER 6 THE APPLICATION OF DATA MINING ON DETERIORATION URGENCY..................................................60 6.1 THE MS SQL 2005 OF DATA MINING.................60 6.1.1 Data proccess..................................61 6.1.2 Association Method.............................64 6.1.3 Clustering Algorithm...........................66 6.1.4 Neural Network.................................69 6.1.5 Accuracy of data mining........................71 6.1.6 The data sensitivity analysis..................72 6.2 THE APPLICATION OF DATA MINING.................74 6.2.1 Input and output model for different algorithm of data mining.............................................74 6.2.2 The factors sensitivity analysis...............82 6.3 ANALYSIS.......................................83 6.3.1 Forecasting Accuracy...........................83 6.3.2 Factors sensitivity analysis...................85 6.3.3 Diagram analysis...............................85 6.4 SUMMARY........................................91 CHAPTER 7 CONCLUSIONS AND SUGGESTIONS...................92 7.1 CONCLUSIONS....................................92 7.2 SUGGESTIONS....................................95 7.3 CONTRIBUTIONS..................................98 APPENDIX I NOTATIONS...................................99 APPENDIX II REFERENCES.................................101 LIST OF FIGURES Figure 2-1 the major items of maintenance works of bridge....................................................6 Figure 2-2 the locations of 20 predefined elements of bridge....................................................9 Figure 3-1 the distribution of condition index...........15 Figure 3-2 the relationship between condition index and age of bridge ................................................15 Figure 3-3 the flow chart of research....................16 Figure 3-4 the damage of bridge..........................23 Figure 3-5 the distribution of bridge length.............24 Figure 3-6 the trend of health condition related to the age of bridge (without weight considered)....................26 Figure 3-7 the trend of health condition related to the age of bridge (with weight considered).......................27 Figure 3-8 the existing performance of each bridge component................................................35 Figure 3-9 the ranking of critical performance index for each component...........................................37 Figure 4-1 the normal health condition of bridge (without weight considered).......................................41 Figure 4-2 the normal health condition of bridge (with weight considered).......................................41 Figure 4-3 the distribution of bridge performance........42 Figure 4-4 the distribution model of health condition....42 Figure 4-5 the diterioration rate of bridge in age between 0 and 13.................................................43 Figure 4-6 the diterioration rate of bridge in age between 13 and 33................................................43 Figure 4-7 the diterioration rate of bridge in age between 33 and 40 ................................................44 Figure 4-8 the performance distribution for each component................................................48 Figure 4-9 the trend of deterioration of bridge based on new table without 0 index................................50 Figure 5-1 the distribtution diagram of normal health condition ................................................53 Figure 5-2 the prediction of life extension for each component ................................................57 Figure 6-1 the basic inspection information of bridge....61 Figure 6-2 the data base of data mining system...........63 Figure 6-3 the model of data mining......................64 Figure 6-4 the probability and importance of the association rules between age of bridge and deterioration urgency..................................................65 Figure 6-5 the dependency of the association rules between age of bridge and deterioration urgency..................66 Figure 6-6 the cluster diagram for different range of bridge age and deterioration urgency value...............68 Figure 6-7 the cluster profiles for different range of bridge age and deterioration urgency value...............68 Figure 6-8 the cluster characteristics for all different range of bridge age and deterioration urgency value......69 Figure 6-9 the maximum value of deterioration urgency for different age (range) of bridge..........................71 Figure 6-10 the accuracy of the output with different analysis algorithm.......................................72 Figure 6-11 the input model with different variables scenario.................................................73 Figure 6-12 the output of probability and accuracy with different input scenario.................................73 Figure 6-13 input model of data mining with factors of bridge age (54Yr, 55Yr5) and U (52Um, 53Uc)..............74 Figure 6-14 the output of data mining with factors of bridge age (54Yr, 55Yr5) and U (52Um, 53Uc)..............75 Figure 6-15 input model of data mining with factors of bridge age (54Yr, 55Yr5) and U (52Um)....................75 Figure 6-16 the output of data mining with factors of bridge age (54Yr, 55Yr5) and U (52Um)....................76 Figure 6-17 input model of data mining with factors of bridge age (54Yr, 55Yr5) and U (53Uc)....................76 Figure 6-18 the output of data mining with factors of bridge age (54Yr, 55Yr5) and U (53Uc)....................77 Figure 6-19 input model of data mining with factors of bridge age (54yr) and U (52um)...........................77 Figure 6-20 the output of data mining with factors of bridge age (54Yr) and U (52 Um)………………………………………………...………………………78 Figure 6-21 input model of data mining with factors of bridge age (55Yr5) and U (52Um)..........................78 Figure 6-22 the output of data mining with factors of bridge age (55Yr5) and U (52Um)..........................79 Figure 6-23 input model of data mining with factors of bridge age (54Yr) and U (53Uc)...........................79 Figure 6-24 the output of data mining with factors of bridge age (54Yr) and U (53 Uc)..........................80 Figure 6-25 input model of data mining with factors of bridge age (55Yr5) and U (53Uc)..........................80 Figure 6-26 the output of data mining with factors of bridge age (55Yr5) and U (53Uc)..........................81 Figure 6-27 input model of data mining with factors of bridge age (55Yr5), type of structure (20~27) and U (53 Uc) .....................................................82 Figure 6-28 the output of data mining with factors of bridge age (55Yr5), type of structure (20~27) and U (53Uc) ..................................................83 Figure 6-29 the diagram of relationship between age of bridge (54Yr and 55Yr5) and degree of deterioration urgency (52Um & 53Uc)............................................86 Figure 6-30 the diagram of relationship between age of bridge (54Yr) and degree of deterioration urgency (52Um & 53Uc)....................................................88 Figure 6-31 the diagram of relationship between age of bridge (55Yr5) and degree of deterioration urgency (52Um & 53Uc)....................................................90 LIST OF TABLES Table 2-1 the comparison of two inspection evaluation methods in Taiwan.........................................7 Table 2-2 the predefined safety components of bridge......9 Table 3-1 predefined bridge elements for inspection......14 Table 3-2 the degree, extent and relevancy (DER) rating for visual inspection........................................14 Table 3-3 the original sample of visual inspection data.....................................................21 Table 3-4 the example of regular inspection report of bridge...................................................22 Table 3-5 the distribution of bridges length.............23 Table 3-6 the screened sample of visual inspection data.....................................................24 Table 3-7 score comparative table of new performance index (pi).....................................................31 Table 3-8 the distribution of new performance index with screened sample of visual inspection data (without weight considering).............................................32 Table 3-9 the distribution of new performance index with screened sample of visual inspection data (with weight considering).............................................32 Table 3-10 the original sample of visual inspection data ....................................................34 Table 3-11 the distribution of the safety performance index for each component in different range of age.............36 Table 3-12 the comparison of the performance index for each component in different range of age......................36 Table 4-1 the maintenance record of the bridge...........39 Table 4-2 the rate of deterioration for the bridge.......44 Table 4-3 the existing deterioration formula for each component of bridge......................................46 Table 4-4 the ranking of deterioration rate for each component of bridge......................................47 Table 4-5 the proposed lifetime without any maintenance..............................................49 Table 4-5 the performance index without 0 index..........51 Table 5-1 the comparison of the allowable performance index to the reliability state.................................56 Table 5-2 the performance recovery of essential maintenance..............................................56 Table 6-1 the scenarios of case study....................74 Table 6-2 the summary table of correct value for each algorithm ................................................84 Table 6-3 the summary table of accuracy percentage for each algorithm ................................................84 Table 6-4 the related years of bridge age (54Yr and 55Yr5) to the degree of deterioration urgency (52Um)............87 Table 6-5 the related years of bridge age (54Yr) to the degree of deterioration urgency (52Um& 53Uc).............89 Table 6-6 the related years of bridge age (55Yr5) to the degree of deterioration urgency (52Um)...................904299759 bytesapplication/pdfen-US檢測績效評估橋梁維護管理劣化安全性資料採礦InspectionPerformanceBridge MaintenanceDeteriorationSafetyData Miningthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/50464/1/ntu-96-D89521017-1.pdf