摘要详情

ID / 提交时间

119 / 2025-05-15 19:03:23

标题

COSRG-LLM: A Design-Phase Intelligent Framework for Operational Support Requirement Generation of Carrier-Based UAVs Using Knowledge Graphs and Large Language Models

关键字

Carrier-based UAV,Knowledge Graph,Large Language Models (LLM)

主题及专题

Special Sessions > SSL 08 New perspective of intelligent detection, diagnosis, prognosis, and maintenance: Generative AI and Industrial Large Model

状态

全文待审

作者

Yi Bi / Beihang University (Beijing University of Aeronautics and Astronautics)

Ruiyang Li / Beihang University；School of Reliability and Systems Engineering

Hao Guo / Beihang University (Beijing University of Aeronautics and Astronautics)

Linhan Guo / Beihang University

Hu Haojie / Beihang University (Beijing University of Aeronautics and Astronautics)

Yuhao Wu / Beihang University (Beijing University of Aeronautics and Astronautics)

Guanlin Liu / Beihang University (Beijing University of Aeronautics and Astronautics)

摘要

Accurate generation of operational support requirements (OSRs) is essential for designing complex defense systems like carrier-based UAVs. Traditional methods often lack semantic traceability and adaptability. This paper presents COSRG-LLM, an intelligent framework combining a domain-specific Knowledge Graph (KG) with a Large Language Model (LLM) to generate structured OSRs. Using a DoDAF-aligned two-stage mapping and three-layer decomposition model, along with a POA-tuning strategy, the system ensures semantic alignment and task adaptability. Validated through an aerial refueling case, COSRG-LLM improves requirement coverage, mapping precision, and reduces design iteration time, offering a scalable method for integrating AI into early equipment design.