Robust optimisation of computationally expensive models using adaptive multi-fidelity emulation

Computationally expensive models are increasingly employed in the design process of engineering products and systems. Robust particular aims to obtain designs that exhibit near-optimal performance low variability under uncertainty. Surrogate often imitate behaviour computational models. Surrogates trained from a reduced number samples model. A crucial component surrogate is quality training set. Problems occur when sampling fails points located an area interest and/or where budget only allows for very limited runs This paper employs Gaussian emulation approach perform efficient single-loop robust optimisation The emulator enhanced propagate input uncertainty output, allowing optimisation. Further, with multi-fidelity data obtained via adaptive maximise set given budget. An illustrative example presented highlight how method works, before it applied two industrial case studies.