建筑結(jié)構(gòu)的地震危險(xiǎn)性和可靠性評(píng)估是一個(gè)不確定的標(biāo)準(zhǔn)。鋼框架結(jié)構(gòu)地震破壞風(fēng)險(xiǎn)突出。通常,地震易損性分析是從概率的角度進(jìn)行的,以保持安全性。初始階段包括標(biāo)定材料、幾何和地震參數(shù)的不確定性。地震易損性分析采用“云”方法完成;該方法需要非線性反應(yīng)時(shí)程分析。對(duì)于以指數(shù)方式增加計(jì)算成本的多個(gè)時(shí)程分析,需要大量的持續(xù)時(shí)間。經(jīng)過(guò)這樣的分析,得到損傷指標(biāo)的整個(gè)過(guò)程是嚴(yán)格的。像人工神經(jīng)網(wǎng)絡(luò)這樣的機(jī)器學(xué)習(xí)技術(shù)正在軟計(jì)算領(lǐng)域出現(xiàn)。以往的研究表明,徑向基函數(shù)神經(jīng)網(wǎng)絡(luò)（RBFNN）能夠在足夠的數(shù)據(jù)條件下預(yù)測(cè)地震損傷。然而,不同結(jié)構(gòu)的神經(jīng)網(wǎng)絡(luò)結(jié)構(gòu)之間缺乏比較。因此,為了更快地預(yù)測(cè)地震易損性曲線,有必要進(jìn)一步尋找最佳的神經(jīng)網(wǎng)絡(luò)及其結(jié)構(gòu)。本文基于機(jī)器學(xué)習(xí)的概念進(jìn)行以下研究。（1）介紹鋼框架結(jié)構(gòu)仿真參數(shù)和性能的選擇和均勻分布。采用均勻設(shè)計(jì)方法對(duì)多因素多層次實(shí)驗(yàn)進(jìn)行均勻抽樣。同時(shí),對(duì)鋼框架的有限元分析進(jìn)行了詳細(xì)地描述。（2）介紹逐步進(jìn)行云分析和繪制地震易損性曲線的方法。結(jié)構(gòu)損傷指標(biāo)采用Park-Ang損傷指數(shù)。從非線性時(shí)程分析中得到的這些數(shù)據(jù)被用于人工網(wǎng)絡(luò)（ANN）的訓(xùn)練、測(cè)試和驗(yàn)證,以預(yù)測(cè)易損性曲線。（3）利用徑向基函數(shù)... 

【文章來(lái)源】：哈爾濱工業(yè)大學(xué)黑龍江省 211工程院校 985工程院校

【文章頁(yè)數(shù)】：81 頁(yè)

【學(xué)位級(jí)別】：碩士

【文章目錄】：
Abstract
中文摘要
Notations
Chapter 1 Introduction
    1.1 Background,objective and significance
        1.1.1 Background
    1.2 Objective and significance
    1.3 Source of the topic
    1.4 Research methodology
    1.5 Development of seismic fragility analysis in structures
        1.5.1 Exploration outside the country
        1.5.2 Exploration inside the country
        1.5.3 Seismic fragility analysis methods and trends
    1.6 Application of neural networks in fragility analysis
    1.7 Main contents of the paper
Chapter 2 Parameter selection,sampling and model setup
    2.1 Introduction
    2.2 Parameter and uncertainty range selection
    2.3 Uniform design method
    2.4 Ground motion selection criteria
    2.5 Selection of ground motions
    2.6 Model setup in ANSYS workbench
    2.7 Summary
Chapter 3 3D-model analysis and fragility curve
    3.1 Introduction
    3.2 Nonlinear Response History Analysis (NRHA)
        3.2.1 Definition
        3.2.2 NRHA vs other analysis
        3.2.3 Importance of NRHA in engineering practice
        3.2.4 Formulation of NRHA
        3.2.5 Application of earthquake acceleration data
    3.3 Theory and application of damage model
        3.3.1 Definition
        3.3.2 Different types of damage index and its use
        3.3.3 Formulation of damage model
    3.4 Fragility curve
        3.4.1 Definition
        3.4.2 Classification of damage index range
        3.4.3 Cloud analysis
        3.4.4 Results
    3.5 Comparison of Fragility Curves with different parameters
    3.6 Problems and challenges
        3.6.1 Obstacles faced during the research
        3.6.2 Challenges in the research
    3.7 Summary
Chapter 4 Fragility prediction using ANN
    4.1 Introduction
    4.2 Theory of artificial neural network
        4.2.1 Definition
        4.2.2 Hard computing vs soft computing
        4.2.3 Architecture and structure of ANN
        4.2.4 Common terms used in ANN
        4.2.5 Classification of artificial neural networks
    4.3 Design of neural network
    4.4 Training and cross-validation
    4.5 Prediction of fragility curve
    4.6 Problems and challenges
    4.7 Summary
Conclusion
References
Acknowledgement
Resume


【參考文獻(xiàn)】：
期刊論文
[1]基于地震觀測(cè)的鋼框架結(jié)構(gòu)有限元模擬與易損性分析[J]. 王飛,康現(xiàn)棟,馬潔美,Kalkan Erol.  防災(zāi)科技學(xué)院學(xué)報(bào). 2018(04)
[2]基于性態(tài)設(shè)計(jì)的鋼筋混凝土結(jié)構(gòu)地震易損性分析[J]. 張進(jìn)國(guó),王洋,徐龍軍.  哈爾濱工程大學(xué)學(xué)報(bào). 2018(10)
[3]基于性能的鋼框架結(jié)構(gòu)地震易損性分析[J]. 李永梅,李玉占,楊博顏.  工程抗震與加固改造. 2017(04)
[4]考慮銹蝕的鋼結(jié)構(gòu)地震易損性分析[J]. 鄭山鎖,田進(jìn),韓言召,徐強(qiáng),孫樂(lè)斌.  地震工程學(xué)報(bào). 2014(01)
[5]不同彎矩增大系數(shù)鋼筋混凝土框架結(jié)構(gòu)地震易損性分析[J]. 張耀庭,馬超,郭宗明,杜曉菊,劉昌芳.  建筑結(jié)構(gòu)學(xué)報(bào). 2014(02)
[6]強(qiáng)柱系數(shù)基于結(jié)構(gòu)易損性的抗震性能評(píng)估[J]. 周靖,蔡健,方小丹.  西北地震學(xué)報(bào). 2007(03)
[7]基于人工神經(jīng)網(wǎng)絡(luò)的土工合成材料加筋擋墻臨界高度預(yù)測(cè)模型(英文)[J]. 周建萍,閆澍旺.  巖土工程學(xué)報(bào). 2002(06)



本文編號(hào)：3551119