數據驅動之鋼筋混凝土構架機率式地震風險評估;Data-driven probabilistic seismic assessment for reinforced concrete frames

NCU Institutional Repository > 工學院 > 土木工程研究所 > 博碩士論文 > Item 987654321/91810

請使用永久網址來引用或連結此文件: http://ir.lib.ncu.edu.tw/handle/987654321/91810

題名:	數據驅動之鋼筋混凝土構架機率式地震風險評估;Data-driven probabilistic seismic assessment for reinforced concrete frames
作者:	李坤展;Lee, Kun-Chan
貢獻者:	土木工程學系
關鍵詞:	OpenSees;機器學習;機率式評估法;增量動力分析;鋼筋混凝土柱;遲滯迴圈;OpenSees;Machine Learning;Probabilistic Assessment Method;Incremental Dynamics Analysis;Reinforced Concrete Columns;Hysteresis Loops
日期:	2023-07-25
上傳時間:	2024-09-19 14:14:32 (UTC+8)
出版者:	國立中央大學
摘要:	台灣社會發展至今，土木建築物仍然持續開發建設，其中鋼筋混凝土構造建築盈千累萬且坐落區域廣泛，因此本研究預期開發一款能有效快速評估鋼筋混凝土構造且提供給使用者簡易操作之軟體，將來可應用於大規模區域性城市監測評估。由於我國位於環太平洋地震帶，板塊運動造成地震頻繁，經由過往地震災損觀察，柱構件在抗震上極其重要，因此本研究藉由國內外學術研究文獻中所收集總共430組鋼筋混凝土柱反覆載重試驗之遲滯迴圈，使用OpenSees模擬分析將實驗數據與模型分析結果進行校準，以主動式學習架構解決繁瑣的過程。此軟體開發藉由Python將OpenSees有限元素軟體導入，針對鋼筋混凝土構造物採用集中塑性模型模擬，並且結合基於樹之機器學習預測柱試體遲滯迴圈非線性參數，以數據驅動進行自動化建模，由門型構架模型進行預測分析成果驗證，供以得知遲滯迴圈以人工智慧預測之參考價值。最後本研究採用機率式建物倒塌耐震評估構架進行非線性增量式動力分析，倒塌判定準則參考PEER-TBI技術報告之整體RC結構倒塌準則，統計回歸成性能等級易損曲線計算建築物之倒塌機率風險，此以性能為導向之評估法可量化震後建物倒榻或受損所造成之災損，包含人員傷亡、修復金額等，提供耐震評估與未來防災政策規劃之需求。 ;Reinforced concrete is one of the most common structures. Therefore, this study is expected to develop a software that can effectively and quickly evaluate reinforced concrete structures and provide users with easy operation, which can be applied to large scale regional urban monitoring and evaluation in the future. Since our country is located in the Pacific Rim seismic zone, where plate motions cause frequent earthquakes, the observation of past earthquake damage indicates that column members are extremely important in seismic resistance. The tedious process was solved by using the OpenSees simulation analysis to calibrate the experimental data with the model analysis results, using an active learning framework. This software was developed by importing OpenSees finite element software in Python, using a centralized plasticity model for reinforced concrete structures, and combining tree-based machine learning to predict the non-linear parameters of the hysteresis loops of the column specimens, using data-driven automated modeling, and validating the results of the predictive analysis by the gantry model, in order to know the reference value of the hysteresis loops predicted by human intelligence. Finally, this study adopts a probabilistic building collapse seismic evaluation framework to conduct non-linear incremental dynamic analysis, and the collapse determination criteria refer to the overall RC structural collapse criteria of PEER-TBI technical report. This performance-oriented assessment method can quantify the damage caused by the collapse or damage of a building after an earthquake, including casualties, repair costs, etc., and provide the needs for seismic evaluation and future disaster prevention policy planning.
顯示於類別:	[土木工程研究所] 博碩士論文

文件中的檔案:

檔案	描述	大小	格式	瀏覽次數
index.html		0Kb	HTML	21	檢視/開啟

在NCUIR中所有的資料項目都受到原著作權保護.

社群 sharing

資料載入中.....