主办:陕西省汽车工程学会
ISSN 1671-7988  CN 61-1394/TH
创刊:1976年

Automobile Applied Technology ›› 2025, Vol. 50 ›› Issue (12): 155-161.DOI: 10.16638/j.cnki.1671-7988.2025.012.027

• Standards·Regulations·Management • Previous Articles    

Research on Micro-Traffic Simulation Optimization Based on Genetic Algorithm

CAI Shuang   

  1. China Merchants Group Vehicle Technology Research Institute Company Limited
  • Published:2025-06-24
  • Contact: CAI Shuang

基于遗传算法的城市微交通仿真优化研究

蔡爽   

  1. 招商局检测车辆技术研究院有限公司
  • 通讯作者: 蔡爽
  • 作者简介:蔡爽(1994-),男,助理工程师,研究方向为汽车碰撞安全性、交通安全仿真研究

Abstract: As a key node of road network, the optimization of signal timing at urban intersections is very important to alleviate traffic congestion. In order to solve the problem that the traditional Webster algorithm has insufficient adaptability in dynamic traffic scenes, this study proposes a single intersection signal timing optimization method based on genetic algorithm (GA). The micro-traffic model of Qingtan West Road intersection is built by SUMO simulation platform, and the real traffic flow characteristics are restored by Krauss car-following model, and several groups of comparative experiments are set up to verify the performance of the algorithm. The results show that the average delay of vehicles at intersections is reduced to 29.84 s (down by 12.31%) and the queue length is shortened to 30.29 m (down by 10.44%), which is significantly better than the optimization effect of Webster algorithm (average delay is 30.30 s/decrease by 10.96%. The queue length is 30.91 m/ decrease by 8.61%). In addition, through floating-point coding and adaptive search mechanism, the genetic algorithm can still achieve the global optimal solution under complex phase constraints, and its traffic capacity is improved to 0.81 saturation, which is 11.0% better than the original scheme. This study proves that genetic algorithm has stronger adaptability and robustness in dynamic traffic signal optimization, which provides a theory for fine management and control of urban intersections.

Key words: simulation optimization; SUMO simulation; Webster algorithm; genetic algorithm

摘要: 城市交叉口作为路网关键节点,其信号配时优化对缓解交通拥堵至关重要。针对传统 Webster 算法在动态交通场景中适应性不足的问题,文章提出基于遗传算法(GA)的单交叉 口信号配时优化方法。通过 SUMO 仿真平台构建青潭西路交叉口微交通模型,采用 Krauss 跟车模型还原真实交通流特性,并设置多组对比实验验证算法性能。研究结果表明,经遗传 算法优化后,交叉口车辆平均延误降低至 29.84 s(降幅 12.31%),排队长度缩短至 30.29 m(降 幅 10.44%),显著优于 Webster 算法的优化效果(平均延误 30.30 s/降幅 10.96%;排队长度 30.91 m/降幅 8.61%)。此外,遗传算法通过浮点数编码与自适应搜索机制,在复杂相位约束 下仍能实现全局最优解,其通行能力提升至 0.81 饱和度,较原方案优化 11.0%。研究证实遗 传算法在动态交通信号优化中具有更强的适应性与鲁棒性,为城市交叉口精细化管控提供了 理论支撑。

关键词: 仿真优化;SUMO 仿真;Webster 算法;遗传算法