大语言模型在消防救援领域的应用现状与展望

消防科学与技术 ›› 2025, Vol. 44 ›› Issue (9): 1311-1319.

大语言模型在消防救援领域的应用现状与展望

韩冰1,2,3, 李大鹏4, 万子敬1,2,3, 李继宝1,2,3, 王澄1,2,3

（1.应急管理部天津消防研究所，天津 300381； 2.工业与公共建筑火灾防控技术应急管理部重点实验室，天津 300381； 3.天津市消防安全技术重点实验室，天津 300381； 4.山东省消防救援总队，山东济南 250102）

收稿日期:2025-03-21 修回日期:2025-04-28 出版日期:2025-09-15 发布日期:2025-09-15
作者简介:韩冰，应急管理部天津消防研究所，助理研究员，主要从事消防救援大模型方面的研究，天津市南开区卫津南路110号，300381，hanbing@tfri.com.cn。
基金资助:
国家重点研发计划项目（2024YFC3017003?4）

Current status and prospects of large language models applied in fire rescue field

Han Bing1,2,3, Li Dapeng4, Wan Zijing1,2,3, Li Jibao1,2,3, Wang Cheng1,2,3

(1. Tianjin Fire Science and Technology Research Institute of MEM, Tianjin 30081, China; 2. Key Laboratory of Fire Protection Technology for Industry and Public Building, Ministry of Emergency Management, Tianjin 300381, China; 3. Tianjin Key Laboratory of Fire Safety Technology, Tianjin 300381, China; 4. Shandong Fire and Rescue Brigade, Jinan Shandong 250102, China)

Received:2025-03-21 Revised:2025-04-28 Online:2025-09-15 Published:2025-09-15

摘要/Abstract

摘要： 为了更好地应对大语言模型对消防救援行业智慧化、数字化转型带来的冲击，首先回顾了大语言模型的发展历程，进而解析了其从通用能力构建到行业适配的发展逻辑；其次，聚焦消防救援场景，系统阐述了大语言模型在消防救援领域的相关技术实现和落地应用现状，基于此，深入剖析大语言模型在行业应用过程中面临的数据可得性、行业适配性等核心挑战；最后，探讨了未来大语言模型在消防救援领域的研究方向与应用前景。

关键词: 消防救援；大语言模型；DeepSeek；自然语言处理；多模态数据；智能体

Abstract: In order to better cope with the impact of large language models on the intelligent and digital transformation of the fire rescue industry, this paper first reviews the development history of large language models, and then analyzes their development logic from general ability construction to industry adaptation. Secondly, focusing on fire rescue scenarios, the system elaborates on the relevant technical implementation and practical application status of large language models in the field of fire rescue. Based on this, it deeply analyzes the core challenges faced by large language models in industry applications, such as data availability and industry adaptability. Finally, the research directions and application prospects of future large language models in the field of fire rescue were discussed.

Key words: fire rescue; large language model; DeepSeek; natural language processing; multimodal data; intelligent agent

韩冰, 李大鹏, 万子敬, 李继宝, 王澄. 大语言模型在消防救援领域的应用现状与展望[J]. 消防科学与技术, 2025, 44(9): 1311-1319.

Han Bing, Li Dapeng4, Wan Zijing, Li Jibao, Wang Cheng. Current status and prospects of large language models applied in fire rescue field[J]. Fire Science and Technology, 2025, 44(9): 1311-1319.