主讲人介绍：Dr. Marco Nie is Professor and Roxelyn and Richard Pepper Family Chair of Civil and Environmental Engineering at Northwestern University. He received his B.S. in Structural Engineering from Tsinghua University, his M.S. from the National University of Singapore, and his Ph.D. from the University of California, Davis. Dr. Nie’s research spans a wide range of topics in transportation systems analysis, transportation economics, and sustainable transportation. He has served on the TRB committees for Transportation Network Modeling and for Traffic Flow Theory and Characteristics. Currently, he is an Area Editor for Transportation Science, and a member of the Editorial Advisory Boards for Transportmetrica-B and Transportation Research Part A and B. His research has been supported by the National Science Foundation, the Transportation Research Board, the U.S. Department of Transportation, the U.S. Department of Energy, and the Illinois Department of Transportation.
摘要：A key challenge in transportation planning is that the collective preferences of the traveling public often diverge from the policies produced by model-driven decision tools. This misalignment frequently results in implementation delays or failures. Here, we investigate whether large language models (LLMs)—noted for their capabilities in reasoning and simulating human decision-making—can help inform and address this alignment problem. We develop a multi-agent simulation in which LLMs, acting as agents representing residents from different communities in a city, participate in a referendum on a set of transit policy proposals. Using chain-of-thought reasoning, LLM agents provide Ranked-Choice or approval-based preferences, which are aggregated using instant-runoff voting (IRV) to model democratic consensus. We implement this simulation framework with both GPT-4o and Claude-3.5-Sonnet, and apply it for Chicago and Houston. Our findings suggest that LLM agents can approximate plausible collective preferences and exhibit sensitivity to local context. At the same time, they display notable deviations from optimization-based benchmarks and behavioral biases that appear specific to the underlying language model. The results underscore both promise and limitations of LLMs as tools for solving the alignment problem in transportation decision-making.
时间：2026年3月16日（周一）上午10:00
地点：西南交通大学犀浦校区交通运输与物流学院417学术报告厅
主办：研究生院交通运输与物流学院
承办：交通运输与物流学院学工组 交通工程系
