arXiv：cs.CL· Zihao Sheng, Pei Li, Zilin Huang, Yen-Jung Chen, Yuhao Luo, Zhengyang Wan, Steven T. Parker, David A. Noyce, Sikai Chen·· 3 小时前
大语言模型辅助交通管理计划编制：WisDOT WisTMP 系统案例研究
Large Language Model-Assisted Preparation of Transportation Management Plans: A Case Study with WisDOT WisTMP System
AI 导读
研究提出一套 LLM 辅助框架，用于自动生成交通管理计划（TMP）内容，并以 WisDOT WisTMP 系统为应用场景。该框架微调多个不同规模的国产开源 LLM 并本地部署以保障数据安全，训练数据由历史 WisTMP 文档 PDF 转成 JSON 问答对构建。实验显示微调显著提升文本生成指标，但模型倾向过度生成策略，难以给出项目专属论证与准确成本估算，且从 7B/8B 扩到 14B 收益有限。
正文
Abstract:Work zones are critical yet hazardous components of transportation infrastructure, requiring carefully designed Transportation Management Plans (TMPs) to ensure safety and mobility. However, TMP preparation remains labor-intensive and heavily dependent on practitioner expertise. This paper proposes a Large Language Model (LLM)-assisted framework to automate TMP content generation, leveraging the WisDOT WisTMP system as the application context. The framework fine-tunes multiple open-source LLMs across different model scales and deploys them locally to ensure data security. To support model training, we construct a domain-specific dataset from historical WisTMP documents by converting PDF files into structured question-answer pairs in JSON format. Experimental results show that fine-tuning significantly improves performance across standard text generation metrics. Further section-wise and strategy-level analyses reveal that, while LLMs achieve strong overall performance, they tend to over-generate strategies and struggle to produce project-specific justifications and accurate cost estimates. In addition, scaling from 7B/8B to 14B yields limited gains. These findings demonstrate the potential of LLMs to improve TMP preparation efficiency while highlighting remaining challenges in LLM-assisted TMP development. The source code and demo videos will be publicly available at this https URL.
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2610.10650 [cs.CL] |
| (or arXiv:2610.10650v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2610.10650 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Zihao Sheng [view email]
[v1]
Wed, 7 Oct 2026 15:51:01 UTC (1,682 KB)
来源：arXiv：cs.CL · arxiv.org