.png){kind=icon}
返回
An intelligent evidence-theory-based structural reliability analysis method based on convolutional neural network model
.png){kind=icon}
PRE
AI
DOI：10.1016/j.cma.2024.116804.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
Combined with the convolutional neural network (CNN) model, an intelligent structural reliability analysis method based on evidence theory is developed to improve the accuracy and efficiency of reliability analysis. Firstly, the uncertainties in engineering structures are described by the Frame of Discernment (FD) and the Basic Probability Assignment (BPA). Secondly, the Optimal Latin Hypercube Design (OLHD) is adopted to obtain the sample focal elements and the Sequential Quadratic Programming (SQP) method is utilized to carry out the extremum analysis of the sample focal elements. Thirdly, the sample focal elements can be reclassified as the belief focal element, the intersect focal element and the failure focal element. Fourthly, the localdensifying method of sample space is applied to ensure the uniformity of sample focal elements and the convolution neural network model is trained by these focal elements. Then, the classifications of the undetermined focal elements could be obtained by the convolution neural network model and the confidence interval can be obtained based on the classification results of all focal elements. Finally, two numerical examples and one engineering application are introduced to investigate the effectiveness of the proposed method.
Keyword:
Evidence theory
Structural reliability analysis
Convolutional neural network
Approximation model
Local-densifying method
期刊
IF:
7.3
论文数:
1.3W
被引数:
5.6W
机构
引用论文
A novel hybrid adaptive Kriging and water cycle algorithm for reliability-based design and optimization strategy: Application in offshore wind turbine monopile用于可靠性设计和优化策略的新型混合自适应Kriging和水循环算法: 在海上风力发电机单桩中的应用
.png){kind=icon}
PREAI
分享收藏
查看更多内容