Microdroplet Generator Surrogate Enhancement and Inverse Design Through an Uncertainty-Guided Framework

The paper introduces a unified framework that combines adaptive surrogate modeling with simulation-driven inverse design for microdroplet generators. High‑fidelity Volume‑of‑Fluid simulations supply training data, while neural‑network ensembles predict flow regimes and droplet diameters; ensemble‑based uncertainty quantification identifies critical parameter regions near regime transitions. A physics‑informed enrichment strategy and Particle Swarm Optimization are used to refine the surrogates and perform constrained inverse design, achieving markedly higher prediction accuracy and reliable identification of physically realizable configurations.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.