Estimation of Response Expectation Bounds under Parametric P-Boxes by Combining Bayesian Global Optimization with Unscented Transform

Abstract

In engineering analysis, propagating parametric probability boxes (p-boxes) remains a challenge because a computationally expensive nested solution scheme is involved. To tackle this challenge, this paper proposes a novel optimization-integration method to propagate parametric probability boxes (p-boxes), mainly focusing on estimating the lower and upper bounds of structural response expectation for linear and moderately nonlinear problems. A model-based optimization scheme, named Bayesian global optimization, is first introduced to explore the space of distribution parameters. Subsequently, an efficient numerical integration method, named unscented transform, is employed to estimate the response expectation with a given set of distribution parameters. Compared with existing optimization-integration methods, the proposed method has three advantages. First, the response expectation bounds are successively estimated, allowing for the reuse of samples generated from the lower-bound estimation in the upper-bound estimation. Second, the approximation error introduced by the numerical integration method is considered. Third, computational efficiency in both the optimization and integration processes is improved. Four applications are investigated to validate the effectiveness of the proposed method, showing its ability to balance computational efficiency and accuracy when evaluating response expectation bounds.