数据与计算发展前沿 ›› 2026, Vol. 8 ›› Issue (4): 30-41.

CSTR: 32002.14.jfdc.CN10-1649/TP.2026.04.003

doi: 10.11871/jfdc.issn.2096-742X.2026.04.003

• • 上一篇    下一篇

基于低秩适配与检索增强融合的望远镜阵列故障诊断专业大模型关键技术

李嘉珊1,2(),邓丽1,*()   

  1. 1 中国科学院国家空间科学中心， 北京 100190
    2 中国科学院大学， 北京 100049
  • 收稿日期:2025-12-24 出版日期:2026-08-20 发布日期:2026-08-21
  • 通讯作者: 邓丽（E-mail: dengli@nssc.ac.cn）
  • 作者简介:李嘉珊，中国科学院国家空间科学中心，硕士研究生。主要研究方向为大语言模型应用、智能数据处理。
    本文承担主要工作为模型整体框架设计、实现、评估与论文撰写。
    LI Jiashan is a master’s student at the National Space Science Center, Chinese Academy of Sciences. Her research interests include applications of large language models and intelligent data processing.
    In this paper, she is mainly responsible for overall model framework design, implementation, evaluation, and manuscript writing.
    E-mail: lijiashan17@mails.ucas.ac.cn|邓丽，中国科学院国家空间科学中心，博士，研究员，博士生导师，CCF会员。主要研究方向为分布式系统信息处理、射电探测技术。
    本文承担主要工作为论文选题指导、实验平台支持与论文修改。
    DENG Li, Ph.D., is a researcher, doctoral supervisor, and member of the CCF at the National Space Science Center, Chinese Academy of Sciences. Her research interests include distributed system information processing and radio detection technology.
    In this paper, she is mainly responsible for.guidance on topic selection, experimental platform support, and manuscript revision.
    E-mail: dengli@nssc.ac.cn

Key Technologies of a Domain-Specific Large Model for Telescope Array Fault Diagnosis Based on the Fusion of Low-Rank Adaptation and Retrieval-Augmented Generation

LI Jiashan1,2(),DENG Li1,*()   

  1. 1 National Space Science Center, Chinese Academy of Sciences, Beijing 100190, China
    2 University of Chinese Academy of Sciences, Beijing 100049, China
  • Received:2025-12-24 Online:2026-08-20 Published:2026-08-21

RichHTML

5

PDF

41

摘要：

【目的】 针对望远镜阵列故障诊断中存在的专业术语晦涩、多源故障信息语义异构以及领域知识稀缺等瓶颈问题，旨在构建一种高精准度与高可靠性的智能诊断模型，以提升大型科学装置的运维效率与系统稳定性。【方法】 本文提出了一种融合低秩适配（LoRA）与检索增强生成（RAG）的故障诊断专业大模型技术框架。该框架以DeepSeek-R1-Qwen3-8B大模型为基座，利用构建的望远镜阵列高质量指令微调数据集，采用LoRA技术实现领域诊断推理能力的高效适配与强化；同时，引入RAG机制建立基于向量数据库的外部知识库，实现故障相关知识片段的实时召回与上下文注入，以增强模型对长尾知识与隐含因果逻辑的覆盖能力。【结果】 以圆环阵太阳射电成像望远镜（DART）为例进行的实验测试表明，该融合模型在BERTScore评估指标上显著优于原始基座模型，其中F1值提升了22.6%。模型在处理复杂故障时展现出更优的诊断准确性、推理稳定性及输出结构化程度。【结论】 LoRA与RAG的协同作用有效解决了通用大模型在专业领域应用中面临的“幻觉”现象与逻辑断层问题。本研究验证了该技术框架的有效性，为望远镜阵列等复杂装备运维场景下的智能故障诊断系统构建提供了可行的技术方案与参考。

关键词: 望远镜阵列, 故障诊断, 低秩适配, 检索增强生成

Abstract:

[Objective] To address the bottleneck issues of obscure specialized terminology, semantic heterogeneity of multi-source fault information, and scarcity of domain knowledge in the fault diagnosis of telescope arrays, this study aims to construct a high-precision and high-reliability intelligent diagnosis model to enhance the operation and maintenance efficiency and system stability of large-scale scientific facilities. [Methods] This paper proposes a technical framework for a domain-specific large model for fault diagnosis based on the fusion of Low-Rank Adaptation (LoRA) and Retrieval-Augmented Generation (RAG). Taking the DeepSeek-R1-Qwen3-8B large model as the backbone, the framework utilizes a constructed high-quality instruction fine-tuning dataset for telescope arrays and employed LoRA technology to efficiently adapt and enhance domain diagnostic reasoning capabilities. Simultaneously, a RAG mechanism is introduced to establish an external knowledge base based on a vector database, enabling the real-time retrieval and context injection of fault-related knowledge fragments to strengthen the model's coverage of long-tail knowledge and implicit causal logic. [Results] Experimental verification using the DAocheng Radio Telescope (DART) as a case study demonstrates that the fused model significantly outperforms the original base model in BERTScore evaluation metrics, with an F1 score improvement of 22.6%. The model exhibits superior diagnostic accuracy, reasoning stability, and output structuredness when handling complex faults. [Conclusion] The synergy of LoRA and RAG effectively addresses the “hallucination” phenomenon and logical gaps faced by general large models in domain-specific applications. This study validates the effectiveness of the proposed technical framework, providing a feasible technical solution and reference for constructing intelligent fault diagnosis systems for radio telescope arrays and other complex equipment operation and maintenance scenarios.

Key words: telescope arrays, fault diagnosis, low-rank adaptation, retrieval-augmented generation