Design of Emergency Command Aided Decision-making System Based on LLM
WU Hao1 DONG Fuan1 WANG Xuefei2 ZHAO Dashuo1 GAO Dong1
1. North Automatic Control Technology Institute, Taiyuan 030006, China; 2. Shanxi Dazhong Electonic Information Industry Group Limited Company, Taiyuan 030027, China
Abstract:Aiming at the needs for rapid recognition and disposal of emergencies in emergency command scenarios, a design of an auxiliary decision-making system based on large Language model(LLM)is proposed. By leveraging the characteristics of natural language understanding, knowledge reasoning and content generation of large models, the efficiency of commanders in handling emergencies and the speed of emergency response can be improved. This system is constructed through the LangChain framework, and adopts technologies such as summary generation, metadata extraction and chain of thought to strengthen the understanding of large models for emergency command scenarios and contexts. It can enhance the ability to grasp the situation of emergencies in the form of comprehensive reports, and meanwhile, it can automatically generate disposal plans to assist commanders in making decisions.
吴昊,董福安,王雪飞,赵大硕,高东. 基于大模型的应急指挥辅助决策系统设计[J]. 指挥与控制学报, 2024, 10(6): 752-759.
WU Hao,DONG Fuan,WANG Xuefei,ZHAO Dashuo,GAO Dong. Design of Emergency Command Aided Decision-making System Based on LLM. Journal of Command and Control, 2024, 10(6): 752-759.