量子計算で流体力学の古典的課題に挑戦(Quantum study feels out questions on fluid flow)

ad

2025-03-25 オークリッジ国立研究所(ORNL)

量子計算で流体力学の古典的課題に挑戦(Quantum study feels out questions on fluid flow)
ORNL researchers used quantum computing to model the unsteady flow of liquids and gases over two parallel plates. Computing time was provided by the Quantum Computing User Program, part of the Oak Ridge Leadership Computing Facility. Credit: Jason Smith/ORNL, U.S. Dept. of Energy

米国オークリッジ国立研究所(ORNL)の研究者らは、量子コンピューティングを活用し、ヘール・ショー流れと呼ばれる流体力学問題に挑みました。この問題は、マイクロ流体工学や地下水流、石油回収などに関連する重要なモデルです。研究では、量子アルゴリズムを使って方程式の解法を模索し、エラー抑制と緩和技術の有効性が示されました。量子計算が流体シミュレーションの革新手段となる可能性を示唆する一方、実用化にはさらなる研究が必要とされています。

<関連情報>

超伝導デバイス上でHarrow-Hassidim-Lloydアルゴリズムを用いたHele-Shaw流の解法: 効率と課題に関する研究
Solving the Hele–Shaw flow using the Harrow–Hassidim–Lloyd algorithm on superconducting devices: A study of efficiency and challenges

Muralikrishnan Gopalakrishnan Meena;Kalyana C. Gottiparthi;Justin G. Lietz;Antigoni Georgiadou;Eduardo Antonio Coello Pérez
Physics of Fluids  Published:October 08 2024
DOI:https://doi.org/10.1063/5.0231929

The development of quantum processors for practical fluid flow problems is a promising yet distant goal. Recent advances in quantum linear solvers have highlighted their potential for classical fluid dynamics. In this study, we evaluate the Harrow–Hassidim–Lloyd (HHL) quantum linear systems algorithm (QLSA) for solving the idealized Hele–Shaw flow. Our focus is on the accuracy and computational cost of the HHL solver, which we find to be sensitive to the condition number, scaling exponentially with problem size. This emphasizes the need for preconditioning to enhance the practical use of QLSAs in fluid flow applications. Moreover, we perform shots-based simulations on quantum simulators and test the HHL solver on superconducting quantum devices, where noise, large circuit depths, and gate errors limit performance. Error suppression and mitigation techniques improve accuracy, suggesting that such fluid flow problems can benchmark noise mitigation efforts. Our findings provide a foundation for future, more complex application of QLSAs in fluid flow simulations.

0106流体工学
ad
ad
Follow
ad
タイトルとURLをコピーしました