Sequential Bayesian optimal experimental design for structural reliability analysis

TL;DR

This paper develops a Bayesian experimental design method for structural reliability analysis, efficiently estimating failure probabilities with minimal costly simulations by combining importance sampling and the unscented transform.

Contribution

It introduces a general hierarchical Bayesian framework for optimal experimental design in structural reliability, addressing both aleatoric and epistemic uncertainties.

Findings

Effective reduction of uncertainty with minimal experiments

Successful application of importance sampling with the unscented transform

Demonstrated efficiency through numerical experiments

Abstract

Structural reliability analysis is concerned with estimation of the probability of a critical event taking place, described by $P(g(\textbf{X}) \leq 0)$ for some $n$-dimensional random variable $\textbf{X}$ and some real-valued function $g$. In many applications the function $g$ is practically unknown, as function evaluation involves time consuming numerical simulation or some other form of experiment that is expensive to perform. The problem we address in this paper is how to optimally design experiments, in a Bayesian decision theoretic fashion, when the goal is to estimate the probability $P(g(\textbf{X}) \leq 0)$ using a minimal amount of resources. As opposed to existing methods that have been proposed for this purpose, we consider a general structural reliability model given in hierarchical form. We therefore introduce a general formulation of the experimental design problem, where we distinguish between the uncertainty related to the random variable $\textbf{X}$ and any additional epistemic uncertainty that we want to reduce through experimentation. The effectiveness of a design strategy is evaluated through a measure of residual uncertainty, and efficient approximation of this quantity is crucial if we want to apply algorithms that search for an optimal strategy. The method we propose is based on importance sampling combined with the unscented transform for epistemic uncertainty propagation. We implement this for the myopic (one-step look ahead) alternative, and demonstrate the effectiveness through a series of numerical experiments.