Multi-stream Physics Hybrid Networks for solving Navier-Stokes equations

Tatjana Protasevich, Mikhail Surmach, Aleksandr Sedykh, Olga Tsurkan, Matvei Anoshin, Vadim Lopatkin, Leonid Fedichkin · 2025

Read Multi-stream Physics Hybrid Networks for solving Navier-Stokes equations by Tatjana Protasevich, Mikhail Surmach, Aleksandr Sedykh, Olga Tsurkan, Matvei Anoshin, Vadim Lopatkin, Leonid Fedichkin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-01891.txt.

Description

Sedykh and co-authors present a hybrid quantum-classical network in two thousand twenty-five for solving fluid flow equations. The Navier-Stokes equations govern fluids, but classic solvers and physics-informed networks both miss parts of the frequency range in a solution. Accuracy and speed suffer together.

Their design splits the solution by frequency, giving each band its own parallel hybrid network that mixes quantum and classical layers. Training simplifies because each stream handles a narrower job. On the classic Kovasznay flow problem, the hybrid beat a matching classical network. Velocity error fell by about thirty-six percent and pressure error by about forty-one percent, with roughly a quarter fewer parameters. Tests covered both data-driven and physics-driven training.

Catalog record

CREATORTatjana Protasevich, Mikhail Surmach, Aleksandr Sedykh, Olga Tsurkan, Matvei Anoshin, Vadim Lopatkin, Leonid Fedichkin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-01891

Browse the full library · About the Tower of Babel