交通运输系统工程与信息 ›› 2025, Vol. 25 ›› Issue (4): 206-218.DOI: 10.16097/j.cnki.1009-6744.2025.04.019

• 系统工程理论与方法 • 上一篇    下一篇

融合仿真与机器学习的交织区通行能力协同估计方法

荣建,吴培佳,高亚聪* ,王益,窦灏   

  1. 广州大学,土木与交通工程学院,广州510006
  • 收稿日期:2025-04-01 修回日期:2025-05-21 接受日期:2025-05-26 出版日期:2025-08-25 发布日期:2025-08-25
  • 作者简介:荣建(1972—),男,四川乐山人,教授。
  • 基金资助:
    广东省哲学社会科学规划青年项目 (GD24YGL31);广州大学研究生创新能力培养项目 (JCCX2024-023)。

A Collaborative Estimation Method for Weaving Area Capacity Integrating Simulation and Machine Learning

RONG Jian, WU Peijia, GAO Yacong*, WANG Yi, DOU Hao   

  1. School of Civil Engineering and Transportation, Guangzhou University, Guangzhou 510006, China
  • Received:2025-04-01 Revised:2025-05-21 Accepted:2025-05-26 Online:2025-08-25 Published:2025-08-25
  • Supported by:
    Youth Program of Philosophy and Social Science Planning of Guangdong Province, China (GD24YGL31);Guangzhou University Graduate Student Innovation Ability Cultivation Program, China (JCCX2024-023)。

摘要: 为克服现有方法在交织流量表征和影响因素量化方面的不足,本文融合微观仿真与机器学习方法,构建从仿真标定和影响因素作用机制分析到通行能力估计的研究框架。提出结合DBSCAN(Density Based Spatial Clustering of Applications with Noise) 聚 类 、信 息 熵 与 遗 传 算 法 的 DIEGA(DBSCAN Information Entropy Genetic Algorithm)仿真标定改进方法;通过仿真实验分析交织区长度( LW )、驶入流量( QRF )、驶出流量(QFR )与通行能力的关联关系;同时,构建基于堆叠策略的通行能力估计模型,并结合SHAP(SHapleyAdditive Explanation)方法剖析各影响因素的作用机制。结果表明:DIEGA标定方法可将交织区各流向延误误差控制在3%以内,较传统遗传算法的收敛速度提升22.2%;在总交织流量相同的情况下,QRF与QFR的不同占比会导致通行能力在约15%范围内波动,且QRF、QFR与LW之间存在非线性耦合关系;基于堆叠策略的随机森林机器学习(ML_RF)模型(R2=0.969)表现最佳,优于其他基线模型;SHAP分析显示,当QRF /QFR占比接近1,且LW的范围为250~350m时,可实现4635~4860pcu·h-1的峰值通行能力。

关键词: 交通工程, 通行能力估计方法, 交通仿真与机器学习, 交织区, 可解释性分析

Abstract: This study integrates microscopic simulation and machine learning to overcome the shortcomings of existing methods in the characterization of weaving flow and the quantification of factors. It constructs a research framework from the simulation calibration, analysis on factor influence, and estimation of traffic capacity. The proposed DIEGA (DBSCAN Information Entropy Genetic Algorithm) method combines DBSCAN (Density-Based Spatial Clustering of Applications with Noise) clustering, information entropy, and genetic algorithms. Simulation experiments explore the relationships among the weaving length ( LW ), on ramp flow ( QRF ), off-ramp flow ( QFR ), and traffic capacity in a weaving area. A stacking-based model of capacity estimation is developed with the SHAP (SHapley Additive Explanation) analysis to reveal how each factor exerts its influence. The results show that DIEGA holds the delay error for each weaving direction below 3%, and converges faster by 22.2% than a traditional genetic algorithm. Under a constant total weaving flow, different proportions of QRF and QFR lead to about 15% fluctuations in capacity. A nonlinear coupling is observed among QRF , QFR , and LW . Among the stacking models, ML_RF ( R2 =0.969) outperforms other stacking approaches and single models. SHAP indicates that when QRF /QFR is near 1 and LW ranges from 250 to 350 meters, a peak capacity of 4635~4860 pcu per hour can be achieved.

Key words: traffic engineering, capacity estimation method, traffic simulation and machine learning, weaving area, interpretability analysis

中图分类号: