Surrogate-Assisted Bounding-Box approach applied to constrained multi-objective optimisation under uncertainty

This paper is devoted to tackling constrained multi-objective optimisation under uncertainty problems. A Surrogate-Assisted Bounding-Box approach (SABBa) formulated here deal with approximated robustness and reliability measures, which can be adaptively refined. defined as a multi-dimensional product of intervals, centred on the estimated objectives constraints, that contains true underlying values. The accuracy these estimations tuned throughout so reach high levels only promising designs, allows quick convergence towards optimal area. In SABBa, this supplemented Surrogate-Assisting (SA) strategy, permits further reduce overall computational cost. adaptive refinement within guided by computation Pareto Optimal Probability (POP) each box. We first assess proposed method several analytical uncertainty-based test-cases respect an priori metamodel in terms probabilistic modified Hausdorff distance set. then applied three engineering applications: design two-bar truss structural mechanics, shape Organic Rankine Cycle turbine blade thermal protection system for atmospheric reentry.