MNO: Multiscale Neural Operator for 3D Computational Fluid Dynamics
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866912925005905920 |
|---|---|
| author | Wang, Qinxuan Wang, Chuang Zhang, Mingyu Sun, Jingwei Yang, Peipei Tang, Shuo Xiang, Shiming |
| author_facet | Wang, Qinxuan Wang, Chuang Zhang, Mingyu Sun, Jingwei Yang, Peipei Tang, Shuo Xiang, Shiming |
| contents | Neural operators have emerged as a powerful data-driven paradigm for solving partial differential equations (PDEs), while their accuracy and scalability are still limited, particularly on irregular domains where fluid flows exhibit rich multiscale structures. In this work, we introduce the Multiscale Neural Operator (MNO), a new architecture for computational fluid dynamics (CFD) on 3D unstructured point clouds. MNO explicitly decomposes information across three scales: a global dimension-shrinkage attention module for long-range dependencies, a local graph attention module for neighborhood-level interactions, and a micro point-wise attention module for fine-grained details. This design preserves multiscale inductive biases while remaining computationally efficient. We evaluate MNO on diverse benchmarks, covering steady-state and unsteady flow scenarios with up to 300k points. Across all tasks, MNO consistently outperforms state-of-the-art baselines, reducing prediction errors by 5% to 50%. The results highlight the importance of explicit multiscale design for neural operators and establish MNO as a scalable framework for learning complex fluid dynamics on irregular domains. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_16071 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MNO: Multiscale Neural Operator for 3D Computational Fluid Dynamics Wang, Qinxuan Wang, Chuang Zhang, Mingyu Sun, Jingwei Yang, Peipei Tang, Shuo Xiang, Shiming Machine Learning Artificial Intelligence Neural operators have emerged as a powerful data-driven paradigm for solving partial differential equations (PDEs), while their accuracy and scalability are still limited, particularly on irregular domains where fluid flows exhibit rich multiscale structures. In this work, we introduce the Multiscale Neural Operator (MNO), a new architecture for computational fluid dynamics (CFD) on 3D unstructured point clouds. MNO explicitly decomposes information across three scales: a global dimension-shrinkage attention module for long-range dependencies, a local graph attention module for neighborhood-level interactions, and a micro point-wise attention module for fine-grained details. This design preserves multiscale inductive biases while remaining computationally efficient. We evaluate MNO on diverse benchmarks, covering steady-state and unsteady flow scenarios with up to 300k points. Across all tasks, MNO consistently outperforms state-of-the-art baselines, reducing prediction errors by 5% to 50%. The results highlight the importance of explicit multiscale design for neural operators and establish MNO as a scalable framework for learning complex fluid dynamics on irregular domains. |
| title | MNO: Multiscale Neural Operator for 3D Computational Fluid Dynamics |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2510.16071 |