An adaptive surrogate model to structural reliability analysis using deep neural network
作者：
Highlights：
• A DNN-based adaptive surrogate model for structural reliability analysis is proposed.
• The performance and limit state functions are evaluated by the surrogate model.
• A threshold is suggested to switch from a globally predicting model to a locally one.
• The paradigm estimates the failure probability with only a small number of samples.
• Six examples are investigated to confirm the reliability of the current methodology.
摘要
•A DNN-based adaptive surrogate model for structural reliability analysis is proposed.•The performance and limit state functions are evaluated by the surrogate model.•A threshold is suggested to switch from a globally predicting model to a locally one.•The paradigm estimates the failure probability with only a small number of samples.•Six examples are investigated to confirm the reliability of the current methodology.
论文关键词：Adaptive surrogate model,Reliability analysis,Monte Carlo Simulation,Deep neural network
论文评审过程：Received 2 April 2020, Revised 11 December 2020, Accepted 14 October 2021, Available online 23 October 2021, Version of Record 3 November 2021.
论文官网地址：https://doi.org/10.1016/j.eswa.2021.116104
