发表机构
Monash University; KTH Royal Institute of Technology; Chinese Academy of Sciences; Zhejiang Dahua Technology; Southeast University（莫纳什大学; 瑞典皇家理工学院; 中国科学院; 浙江大华技术股份有限公司; 东南大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文提出EvoSignal，一种LLM引导的进化框架，用于模块化交通信号控制程序的设计，通过性能反馈优化策略，在真实网络模拟中显著降低等待时间，并实现跨场景迁移。
AI 中文摘要
有效的交通信号控制（TSC）需要能够响应不断变化的交通需求和网络状况，同时满足不同控制目标的策略。然而，调整现有策略通常涉及重复的人工设计和调整，使得难以系统性地探索目标网络的更好控制规则。大型语言模型（LLMs）可以自动化这一过程，但直接使用它们来选择信号相位会将决策规则嵌入黑盒模型，并产生重复的推理成本和延迟。本文将TSC表述为模块化程序设计问题，并提出EvoSignal，一个利用交通知识和性能反馈的LLM引导进化框架。模块化表示将交通特征提取、局部相位优先级排序和可选的基于网络的优先级调整分开。从几个既定策略开始，EvoSignal通过拥堵和信号操作的反馈改进程序，保留具有不同性能权衡的策略。所得程序无需在线LLM推理即可运行。在两个真实道路网络上的五个场景中的模拟实验表明，所选默认EvoSignal程序相对于每个场景中20个传统、基于强化学习和基于LLM的基线所达到的最低等待时间，减少了16.8%至49.2%的等待时间。一个优先考虑旅行时间和队列长度的程序在搜索场景中在所有三个指标上优于所有20个基线，并且在未经修改转移到其他四个场景时，在每个指标上保持前三名。这些发现支持可检查控制程序的自动化设计，这些程序可跨评估的道路网络和交通条件迁移。代码可在以下网址获取：此https URL。
英文摘要
Effective traffic signal control (TSC) requires policies that respond to changing traffic demand and network conditions while meeting different control objectives. However, adapting existing strategies often involves repeated manual design and adjustment, making it difficult to systematically explore better control rules for a target network. Large language models (LLMs) can automate this process, but directly using them to select signal phases leaves decision rules embedded in black-box models and incurs recurring inference costs and latency. This paper formulates TSC as a modular program design problem and proposes EvoSignal, an LLM-guided evolutionary framework using traffic knowledge and performance feedback. The modular representation separates traffic feature extraction, local phase prioritization, and optional network-based priority adjustment. Starting from several established strategies, EvoSignal improves programs through feedback on congestion and signal operation, retaining strategies with different performance trade-offs. The resulting programs operate without online LLM inference. Simulation experiments across five scenarios on two real-world road networks show that the selected default EvoSignal program reduces waiting time by 16.8--49.2\% relative to the lowest waiting time achieved by the 20 conventional, reinforcement learning-based, and LLM-based baselines in each scenario. A program prioritizing travel time and queue length outperforms all 20 baselines on all three metrics in the search scenario and remains among the top three on each metric when transferred unchanged to the other four scenarios. These findings support automated design of inspectable control programs that transfer across the evaluated road networks and traffic demands.Code is available at https://github.com/georgewanglz2019/EvoSignal.
