An Efficient Parallelized Adaptive Learning Framework for Small Failure Probability Analysis

doi:10.1016/j.apm.2026.116785

An Efficient Parallelized Adaptive Learning Framework for Small Failure Probability Analysis

2026-01-21

0

PRE

AI

DOI：10.1016/j.apm.2026.116785

Original

Original request for help

Save

Abstract

En 中文

• An effective parallelized infilling criterion is designed using influence factors; • An error-based function allocation strategy is introduced for sample selection; • A hybrid convergence criterion is proposed to terminate adaptive learning; • A MCMC-IS sampling method is implemented for rare failure event analysis; • The proposed method demonstrates high efficiency and accuracy in four case studies.

An Efficient Parallelized Adaptive Learning Framework for Small Failure Probability Analysis

Abstract

Journal

Organization