跳到正文
arXiv cs.AI· S{\o}ren Taverniers, Eric Darve, Hamdi A. Tchelepi·· 1 小时前AI 评分36

以架构保障可靠性:用 Agentic 工作流从自然语言实现经校验的多物理场仿真

Reliability by Architecture: Agentic Workflows for Validated Multi-Physics Simulation from Natural Language

AI 导读

一项研究提出 agentic 工作流,用 LLM 结合确定性例程协调 GEOS 多物理场仿真器的建模、执行与校验。在两个流动问题的 80 次尝试中,接地处理使结构有效输入比例从 5/20 提升至 20/20,但仅配合完整的规则化输入准备才能达到 20/20 验收;相同输入下局部修复可修正 3 个植入缺陷并保留有效输入。

正文 · 原文

View PDF HTML (experimental)

Abstract:Preparing, executing, and validating strongly coupled multi-physics simulations require expert time. Agentic workflows use large language models (LLMs) to coordinate them from natural-language descriptions, but reliability requires checks of physical consistency and fidelity to the problem description. Our agentic harness implements explicit task allocation: LLMs interpret user requests and simulator diagnostics, generate input components not covered by predefined construction rules, and propose repairs, while deterministic routines construct supported components and check inputs, edits, and results. We evaluate CO2-storage problems using the multi-physics simulator GEOS, from compositional flow to flow-poromechanics with frictional fault contact. In 80 attempts on two flow problems, grounding raises the proportion of structurally valid inputs from 5/20 to 20/20, but acceptance reaches 20/20 only with complete rule-based input preparation. With identical inputs, local repairs correct three seeded defects while preserving valid inputs. Supporting tests identify simulations passing physics checks despite differing from the problem description or reference results.
Subjects: Computational Engineering, Finance, and Science (cs.CE); Artificial Intelligence (cs.AI); Computational Physics (physics.comp-ph)
Cite as: arXiv:2611.23473 [cs.CE]
  (or arXiv:2611.23473v1 [cs.CE] for this version)
  https://doi.org/10.48550/arXiv.2611.23473

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Søren Taverniers [view email]
[v1] Tue, 17 Nov 2026 20:25:26 UTC (64 KB)

来源:arXiv cs.AI · arxiv.org