A GRU-based ensemble learning method for time-variant uncertain structural response analysis

被引:28
作者
Zhang, Kun [1 ]
Chen, Ning [1 ]
Liu, Jian [1 ]
Beer, Michael [2 ,3 ,4 ]
机构
[1] Hunan Univ, State Key Lab Adv Design & Mfg Vehicle Body, Changsha, Peoples R China
[2] Leibniz Univ Hannover, Inst Risk & Reliabil, Callinstr 34, Hannover, Germany
[3] Univ Liverpool, Inst Risk & Uncertainty, Peach St, Liverpool L69 7ZF, England
[4] Tongji Univ, Int Joint Res Ctr Engn Reliabil & Stochast Mech, Shanghai 200092, Peoples R China
基金
中国国家自然科学基金;
关键词
GRU; Time-variant response; Active learning; Gaussian process; Ensemble learning; FINITE-ELEMENT; REDUCED BASIS; INTERVAL; RELIABILITY; SIMULATION; EXPANSION; SYSTEMS; DESIGN; MODELS;
D O I
10.1016/j.cma.2021.114516
中图分类号
T [工业技术];
学科分类号
08 ;
摘要
Owing to the influence of manufacturing and assembly errors, material performance degradation, external loads and unpredictability of the environment during service, structural response analysis should consider the time-invariant uncertainties and time-variant uncertainties simultaneously. In this paper, a mixed uncertainty model with random variable and stochastic process is adopted to handle this issue. A time-variant uncertain structural response analysis method is proposed based on recurrent neural network using gated recurrent units (GRU) combined with ensemble learning. In the proposed method, by performing Latin hypercube sampling (LHS) of random variables, multiple GRU networks can be trained to estimate the time-variant system response under fixed random variables. During the process of training GRU models, an active learning strategy is developed and applied to improve model accuracy and reduce training samples. On this basis, a set of augmented data is generated using the trained GRU models. Then the mapping relationship between random variables and structural responses through the Gaussian process (GP) regression is built accordingly. Eventually, the global surrogate model of time-variant uncertain structural response can be obtained by integrating the GRU networks and the GP models. Two numerical examples are used to demonstrate the effectiveness and accuracy of the proposed method. The results indicate that the proposed method can effectively calculate the expectation and standard deviation of the system response under the mixed uncertainty model with random variables and stochastic processes. In addition, it has higher computational efficiency under the premise of ensuring the calculation accuracy. (c) 2021 Elsevier B.V. All rights reserved.
引用
收藏
页数:20
相关论文
共 40 条
  • [2] [Anonymous], 2018, Recent advances in deep learning: An overview.
  • [3] Box J.F., 1987, Statistical Science, V2, P45, DOI [DOI 10.1214/SS/1177013437, 10.1214/ss/1177013437]
  • [4] A fast Monte-Carlo method with a reduced basis of control variates applied to uncertainty propagation and Bayesian estimation
    Boyaval, Sebastien
    [J]. COMPUTER METHODS IN APPLIED MECHANICS AND ENGINEERING, 2012, 241 : 190 - 205
  • [5] Boyaval S, 2010, COMMUN MATH SCI, V8, P735
  • [6] Hybrid interval and random analysis for structural-acoustic systems including periodical composites and multi-scale bounded hybrid uncertain parameters
    Chen, Ning
    Xia, Siyuan
    Yu, Dejie
    Liu, Jian
    Beer, Michael
    [J]. MECHANICAL SYSTEMS AND SIGNAL PROCESSING, 2019, 115 : 524 - 544
  • [7] Karhunen-Loeve expansion for multi-correlated stochastic processes
    Cho, H.
    Venturi, D.
    Karniadakis, G. E.
    [J]. PROBABILISTIC ENGINEERING MECHANICS, 2013, 34 : 157 - 167
  • [8] Cho K., 2014, P C EMP METH NAT LAN, P1724
  • [9] Using recurrent neural network models for early detection of heart failure onset
    Choi, Edward
    Schuetz, Andy
    Stewart, Walter F.
    Sun, Jimeng
    [J]. JOURNAL OF THE AMERICAN MEDICAL INFORMATICS ASSOCIATION, 2017, 24 (02) : 361 - 370
  • [10] Review and application of Artificial Neural Networks models in reliability analysis of steel structures
    Chojaczyk, A. A.
    Teixeira, A. P.
    Neves, L. C.
    Cardoso, J. B.
    Guedes Soares, C.
    [J]. STRUCTURAL SAFETY, 2015, 52 : 78 - 89