New AI framework CoFINN improves fluid flow prediction by embedding physics

Researchers have developed CoFINN (Conservation Flux Informed Neural Networks), a novel deep learning framework designed to predict fluid flow fields governed by conservation laws. Unlike standard CNNs that focus on pixel similarity, CoFINN integrates finite-volume conservation physics directly into its training process. This approach treats CNN outputs as structured grids and enforces conservation through numerical flux calculations, improving accuracy in predicting aerodynamic forces, particularly in low-data scenarios. The framework has demonstrated a reduction in drag prediction error by up to 34% at extreme angles of attack. AI

IMPACT This framework could enhance the accuracy and efficiency of AI models used in scientific simulations, particularly in fluid dynamics.

RANK_REASON This is a research paper detailing a new AI framework for a specific scientific domain. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• Adnan Harun Dogan
• CNN
• CoFINN
• Conservation Flux Informed Neural Networks

• paper
• infra

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →