arXiv cs.AI· Guoyang Shi, Hongwen Li, Yiding Yu, Zhuang Fu, Siqi Ding, Zhengyuan Chen, JianGuo Chen, Hanyue Zhao, Tianyuan Liu·· 2 小时前AI 评分31
多模态智能体实现对非平稳响应科学仪器的自适应控制
Adaptive Control of Scientific Instruments with Nonstationary Responses through Multimodal Agents
AI 导读
论文提出一种物理可解释的多模态智能体框架，用于控制响应非平稳的科学仪器，以磁约束聚变装置中性束注入（NBI）离子源的逐炮调参为验证场景。该框架由确定性算法提取符号物理量与视觉证据、双层记忆组织历史轨迹、多模态大语言模型生成候选动作并经独立可行性检查执行。
正文 · 原文
Abstract:Controlling scientific instruments requires parameters to be adjusted progressively in response to feedback from successive operations. When unobserved operating conditions evolve with operational history, responses become nonstationary, and a static mapping from actions to responses becomes inadequate for decisions across shots. Shot-to-shot tuning of neutral beam injection (NBI) ion sources in magnetic confinement fusion devices is representative of this class of problems: the conditions of electrode surfaces and vacuum are difficult to observe directly, discharge outcomes vary between shots, and experiments are costly and carry operational risks. We propose a physically grounded multimodal agent with three independently auditable components. Deterministic algorithms extract symbolic physical quantities and visual evidence from multichannel waveforms; a memory with two layers organizes trajectories from the current task, experience from earlier tasks, and procedural knowledge, from which a multimodal large language model (MLLM) generates candidate actions; and an independent action feasibility check enforces action constraints before execution. In a strictly chronological replay of 455 shots from four EXL-50U platforms, physical symbols alone outperformed waveform images alone by 0.21-0.32 in macro-F1 for quality classification across all four evaluated models. Retrieved context increased the mean agreement with historical action directions by 22.4-42.7 percentage points, whereas the contribution of visual evidence varied across models. Under supervised operation on five ion sources, the system achieved a beam-extraction rate of 86.80% across 409 executed shots. These results demonstrate the engineering feasibility of the framework for supervised sequential tuning on a physical scientific instrument.
| Subjects: | Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2610.32370 [cs.AI] |
| (or arXiv:2610.32370v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.32370 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Guoyang Shi [view email]
[v1]
Fri, 23 Oct 2026 08:59:43 UTC (6,574 KB)
来源：arXiv cs.AI · arxiv.org