Reliability Analysis of Structures by Active Learning Enhanced Sparse Bayesian Regression

Adaptive sampling near a limit state is important for metamodeling-based reliability analysis of structures involving an implicit function. Active learning based on the posterior mean and standard deviation provided by chosen metamodel widely used such adaptive sampling. Most studies active learning-based estimation methods use Kriging approach, which provides prediction along with its variance. As sparse Bayesian regression also deviation. Due to sparsity involved in learning, it expected be computationally faster than approach. Motivated this, learning-enhanced sampling-based explored present study analysis. In doing so, polynomial basis functions, do not involve free parameters, are avoid expensive parameter tuning. The convergence proposed approach attained stabilization 10 consecutive failure estimates. effectiveness illustrated numerically five examples.