Journal of Engineering Research

The rapid development of AI, especially the breakthroughs of LLMs in natural language processing, has opened up new technical avenues for the intelligent generation and precise retrieval of electronic archives. This paper, using the Ruicang Highway Project as a basis, explores the intelligent generation and precise retrieval of traffic engineering electronic archives based on LLMs. By building a domain - specific knowledge base and refining semantic understanding algorithms, it aims to automate and ensure compliance in archive content generation while improving retrieval efficiency. These efforts enhance archive management and reduce carbon emissions, thereby supporting the digital management of traffic engineering throughout its entire life cycle. The study focuses on addressing key issues such as archive standardization, technological adaptability, and data security, with the goal of advancing the green and smart transportation system.

Large language models; Traffic engineering; Electronic archives; Intelligent generation; Precise retrieval