Machine learning aided stochastic elastoplastic analysis

Abstract

The stochastic elastoplastic analysis is investigated for structures under plane stress/strain conditions. A novel uncertain nonlinear analysis framework, namely the machine leaning aided stochastic elastoplastic analysis (MLA-SEPA), is presented herein via finite element method (FEM). The proposed MLA-SEPA is a favourable alternative to determine structural reliability when full-scale testing is not achievable, thus leading to significant eliminations of manpower and computational efforts spent in practical engineering applications. Within the MLA-SEPA framework, an extended support vector regression (X-SVR) approach is introduced and then incorporated for the subsequent uncertainty quantification. By successfully establishing the governing relationship between the uncertain system parameters and any concerned structural output, a comprehensive probabilistic profile including means, standard deviations, probability density functions (PDFs), and cumulative distribution functions (CDFs) of the structural output can be effectively established through a sampling scheme. Consequently, the nonlinear performance of the structure against both serviceability and strength limit states can be effectively investigated with the consideration of various system uncertainties. Three numerical examples are thoroughly investigated to illustrate the accuracy, applicability and effectiveness of the proposed MLA-SEPA approach.