Research and Application of a Multi-Modal Information Scheduling Method Based on Large Language Models in Intelligent Transportation

As intelligent transportation systems evolve, traffic management faces increasing challenges in processing vast amounts of heterogeneous data and supporting complex decision-making processes. To address these challenges, a multi-modal information scheduling method based on large language models is proposed, enabling efficient parsing of natural language inputs, cross-database queries, and automated scheduling through functionalities such as intelligent POI scheduling, associated information query, and precise camera scheduling. The findings demonstrate that this approach significantly enhances decision-support capabilities in complex scenarios and offers valuable insights for applications within the intelligent transportation domain.