系统仿真学报 ›› 2025, Vol. 37 ›› Issue (6): 1343-1351.doi: 10.16182/j.issn1004731x.joss.24-1246

• 新质交通系统建模与仿真 •    

基于LSTM-GNN的畸形交叉口自适应信号控制仿真研究

陈坤, 陈亮, 谢济铭, 刘丰博, 陈泰熊, 位路宽   

  1. 昆明理工大学 交通工程学院,云南 昆明 650093
  • 收稿日期:2024-11-11 修回日期:2025-03-04 出版日期:2025-06-20 发布日期:2025-06-18
  • 通讯作者: 陈亮
  • 第一作者简介:陈坤(2000-),男,硕士生,研究方向为交通安全与仿真。

Simulation Study on Adaptive Signal Control of Deformed Intersection Based on LSTM-GNN

Chen Kun, Chen Liang, Xie Jiming, Liu Fengbo, Chen Taixiong, Wei Lukuan   

  1. School of Traffic Engineering, Kunming University of Science and Technology, Kunming 650093, China
  • Received:2024-11-11 Revised:2025-03-04 Online:2025-06-20 Published:2025-06-18
  • Contact: Chen Liang

摘要:

针对畸形交叉口交通拥堵情况,设计了一种基于深度学习的改进型自适应交通信号控制方案,融合了LSTM与GNN在畸形交叉口的自适应信号控制。LSTM捕捉时间序列交通数据之间的依赖性,GNN构建车道间的空间交互模型。通过整合时间和空间维度的信息,该模型能够依据实时交通状况动态调整信号灯的相位时长。结果表明:LSTM-GNN自适应控制方案相比传统固定信号控制提高了约17.3%的整体通过效率。

关键词: 深度学习, LSTM, GNN, 交通信号控制, 畸形交叉口, 自适应控制, 交通流优化, 时空依赖性

Abstract:

Aiming at the traffic congestion at deformed intersections, an improved adaptive traffic signal control scheme based on deep learning is designed, the scheme integrates the adaptive signal control of LSTM and GNN at deformed intersections. LSTM is used to capture the dependence between time series traffic data, while GNN is used to construct a spatial interaction model between lanes. By integrating the information of time and space dimensions, the model can dynamically adjust the phase duration of signal lights according to real-time traffic conditions. The results indicate that the LSTM-GNN adaptive control scheme improves overall traffic throughput efficiency by approximately 17.3% compared to traditional fixed-signal control.

Key words: deep learning, LSTM, GNN, traffic signal control, deformed intersection, adaptive control, traffic flow optimization, time-space dependence

中图分类号: