Reproduce “A Multi-Fidelity Methodology for Reduced Order Models with High-Dimensional Inputs” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Bilal Mufti, Christian Perron, Dimitri N. Mavris
A multi-fidelity reduced order modeling framework, MF-PCAS, was developed to address the curse of dimensionality in aerospace design by simultaneously reducing input and output dimensions. The framework integrates manifold alignment and model-based active subspaces, demonstrating improved prediction accuracy and computational cost-efficiency on airfoil and wing designs with up to 60 variables.
Key insights
- The MF-PCAS framework simultaneously reduces both input and output dimensions by integrating active subspace methods and manifold alignment, addressing a limitation of previous ROM techniques.
- A model-based active subspace approach is employed, eliminating the need for direct gradient information from high-fidelity simulations and broadening the method's applicability to black-box codes.
- Multi-fidelity data is leveraged through manifold alignment via Procrustes analysis, ensuring compatible latent space representations for effective data fusion across different fidelity levels.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.