A Monte Carlo Simulation Approach in Non-linear Structural Dynamics Using Convolutional Neural Networks

被引:9
|
作者
Bamer, Franz [1 ]
Thaler, Denny [1 ]
Stoffel, Marcus [1 ]
Markert, Bernd [1 ]
机构
[1] Rhein Westfal TH Aachen, Inst Gen Mech, Aachen, Germany
关键词
Monte Carlo method; non-linear structural mechanics; elastoplastic structure; convolutional neural networks; machine learning; earthquake engineering; probability of failure; STRATEGY;
D O I
10.3389/fbuil.2021.679488
中图分类号
TU [建筑科学];
学科分类号
0813 ;
摘要
The evaluation of the structural response statistics constitutes one of the principal tasks in engineering. However, in the tail region near structural failure, engineering structures behave highly non-linear, making an analytic or closed form of the response statistics difficult or even impossible. Evaluating a series of computer experiments, the Monte Carlo method has been proven a useful tool to provide an unbiased estimate of the response statistics. Naturally, we want structural failure to happen very rarely. Unfortunately, this leads to a disproportionately high number of Monte Carlo samples to be evaluated to ensure an estimation with high confidence for small probabilities. Thus, in this paper, we present a new Monte Carlo simulation method enhanced by a convolutional neural network. The sample-set used for this Monte Carlo approach is provided by artificially generating site-dependent ground motion time histories using a non-linear Kanai-Tajimi filter. Compared to several state-of-the-art studies, the convolutional neural network learns to extract the relevant input features and the structural response behavior autonomously from the entire time histories instead of learning from a set of hand-chosen intensity inputs. Training the neural network based on a chosen input sample set develops a meta-model that is then used as a meta-model to predict the response of the total Monte Carlo sample set. This paper presents two convolutional neural network-enhanced strategies that allow for a practical design approach of ground motion excited structures. The first strategy enables for an accurate response prediction around the mean of the distribution. It is, therefore, useful regarding structural serviceability. The second strategy enables for an accurate prediction around the tail end of the distribution. It is, therefore, beneficial for the prediction of the probability of failure.
引用
收藏
页数:10
相关论文
共 50 条
  • [1] A conditionally linearized Monte Carlo filter in non-linear structural dynamics
    Sajeeb, R.
    Manohar, C. S.
    Roy, D.
    [J]. INTERNATIONAL JOURNAL OF NON-LINEAR MECHANICS, 2009, 44 (07) : 776 - 790
  • [2] Structural reliability analysis using Monte Carlo simulation and neural networks
    Cardoso, Joao B.
    de Almeida, Joao R.
    Dias, Jose M.
    Coelho, Pedro G.
    [J]. ADVANCES IN ENGINEERING SOFTWARE, 2008, 39 (06) : 505 - 513
  • [3] ECG Arrhythmia Classification Using Non-Linear Features and Convolutional Neural Networks
    Cajas, Sebastian
    Astaiza, Pedro
    Garcia-Chicangana, David Santiago
    Segura, Camilo
    Lopez, Diego M.
    [J]. 2020 COMPUTING IN CARDIOLOGY, 2020,
  • [4] Non-linear failure rate: A Bayes study using Hamiltonian Monte Carlo simulation
    Thach, Tien T.
    Bris, Radim
    Volf, Petr
    Coolen, Frank P. A.
    [J]. INTERNATIONAL JOURNAL OF APPROXIMATE REASONING, 2020, 123 : 55 - 76
  • [5] Reliability-based structural optimization using neural networks and Monte Carlo simulation
    Papadrakakis, M
    Lagaros, ND
    [J]. COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING, 2002, 191 (32) : 3491 - 3507
  • [6] Using Deep Convolutional Neural Networks in Monte Carlo Tree Search
    Graf, Tobias
    Platzner, Marco
    [J]. COMPUTERS AND GAMES, CG 2016, 2016, 10068 : 11 - 21
  • [7] MONTE-CARLO SIMULATION OF THE NON-LINEAR FOKKER - PLANCK EQUATION
    JONES, RD
    LEE, K
    LEE, YC
    SAMEC, TK
    [J]. BULLETIN OF THE AMERICAN PHYSICAL SOCIETY, 1980, 25 (08): : 985 - 985
  • [8] Interacting dipole charges in non-linear dielectrics: A Monte Carlo simulation
    Kliem, H
    Farag, N
    [J]. IEEE 1997 ANNUAL REPORT - CONFERENCE ON ELECTRICAL INSULATION AND DIELECTRIC PHENOMENA, VOLS I AND II, 1997, : 11 - 14
  • [9] Analysis of non-linear activation functions for classification tasks using convolutional neural networks
    Dureja, Aman
    Pahwa, Payal
    [J]. Recent Patents on Computer Science, 2019, 12 (03) : 156 - 161