Saved in:
Bibliographic Details
Main Authors: Du, Peimian, Liu, Jiabin, Jin, Xiaowei, Zuo, Wangmeng, Li, Hui
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.11578
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912895407751168
author Du, Peimian
Liu, Jiabin
Jin, Xiaowei
Zuo, Wangmeng
Li, Hui
author_facet Du, Peimian
Liu, Jiabin
Jin, Xiaowei
Zuo, Wangmeng
Li, Hui
contents This research confronts the challenge of substantial physical equation discrepancies encountered in the generation of spatiotemporal physical fields through data-driven trained models. A spatiotemporal physical field generation model, named HMT-PF, is developed based on the hybrid Mamba-Transformer architecture, incorporating unstructured grid information as input. A fine-tuning block, enhanced with physical information, is introduced to effectively reduce the physical equation discrepancies. The physical equation residuals are computed through a point query mechanism for efficient gradient evaluation, then encoded into latent space for refinement. The fine-tuning process employs a self-supervised learning approach to achieve physical consistency while maintaining essential field characteristics. Results show that the hybrid Mamba-Transformer model achieves good performance in generating spatiotemporal fields, while the physics-informed fine-tuning mechanism further reduces significant physical errors effectively. A MSE-R evaluation method is developed to assess the accuracy and realism of physical field generation.
format Preprint
id arxiv_https___arxiv_org_abs_2505_11578
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Spatiotemporal Field Generation Based on Hybrid Mamba-Transformer with Physics-informed Fine-tuning
Du, Peimian
Liu, Jiabin
Jin, Xiaowei
Zuo, Wangmeng
Li, Hui
Machine Learning
Artificial Intelligence
Computational Physics
This research confronts the challenge of substantial physical equation discrepancies encountered in the generation of spatiotemporal physical fields through data-driven trained models. A spatiotemporal physical field generation model, named HMT-PF, is developed based on the hybrid Mamba-Transformer architecture, incorporating unstructured grid information as input. A fine-tuning block, enhanced with physical information, is introduced to effectively reduce the physical equation discrepancies. The physical equation residuals are computed through a point query mechanism for efficient gradient evaluation, then encoded into latent space for refinement. The fine-tuning process employs a self-supervised learning approach to achieve physical consistency while maintaining essential field characteristics. Results show that the hybrid Mamba-Transformer model achieves good performance in generating spatiotemporal fields, while the physics-informed fine-tuning mechanism further reduces significant physical errors effectively. A MSE-R evaluation method is developed to assess the accuracy and realism of physical field generation.
title Spatiotemporal Field Generation Based on Hybrid Mamba-Transformer with Physics-informed Fine-tuning
topic Machine Learning
Artificial Intelligence
Computational Physics
url https://arxiv.org/abs/2505.11578