中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊

重庆交通大学学报(自然科学版) ›› 2022, Vol. 41 ›› Issue (12): 41-47.DOI: 10.3969/j.issn.1674-0696.2022.12.06

• 交通+大数据人工智能 • 上一篇    下一篇

基于投影寻踪法的城市道路横断面评价方法

张志清,宋欣,付金秀   

  1. (北京工业大学 城市交通学院 北京市交通工程重点实验室,北京 100124)
  • 收稿日期:2021-03-26 修回日期:2021-06-15 发布日期:2023-01-16
  • 作者简介:张志清(1965—),男,内蒙古卓资人,教授,博士,主要从事现代道路交通安全设计方法的研究。E-mail:zhangzhiqing@bjut.edu.cn 通信作者:宋欣(1996—),男,河北石家庄人,硕士研究生,主要从事现代道路规划与设计方面的研究。E-mail:1165030752@qq.com
  • 基金资助:
    国家自然科学基金项目(51578027)

Evaluation Method of Urban Road Cross-Section Based on Projection Pursuit Method

ZHANG Zhiqing, SONG Xin, FU Jinxiu   

  1. (Beijing Key Laboratory of Traffic Engineering, College of Metropolitan Transportation, Beijing University of Technology, Beijing 100124, China)
  • Received:2021-03-26 Revised:2021-06-15 Published:2023-01-16

摘要: 为减少现有评价方法中主观因素对城市道路设计阶段横断面方案选型的影响,从交通运行效率、交通安全以及资源环境三个方面构建了城市道路横断面选型评价指标体系。基于投影寻踪基本原理,通过构造投影指标函数,采用实数编码加速遗传算法完成最优投影方向寻优计算,确定了评价指标的最优权重,构建了横断面评价的实数编码加速遗传算法-投影寻踪模型,并应用于工程实例与其他评价方法对比。结果表明:三种典型横断面形式的投影寻踪评价结果分别为0.571、-0.145、0.465,模糊综合评价结果分别为0.544、0.427、0.537,通过对比两个评价方法发现,投影寻踪法可将横断面设计因素产异性在结果上进行放大,从而减少主观因素对评价结果的影响,使城市道路横断面设计方案选型更加科学合理。

关键词: 道路工程;道路横断面;投影寻踪法;实数编码加速遗传算法;最优投影向量

Abstract: In order to reduce the impact of subjective factors in the existing evaluation methods on the selection of cross section schemes in the design stage of urban roads, an evaluation index system for the selection of cross section schemes of urban roads was established from three aspects, such as traffic operation efficiency, traffic safety and resources and environment. Based on the basic principle of the projection pursuit, the projection index function was constructed, and the real number coding accelerated genetic algorithm was used to complete the optimization calculation of the optimal projection direction. The optimal weights of evaluation indexes were determined finally. A real number coding accelerated genetic algorithm -projection pursuit model for cross section evaluation was constructed and applied to a project case to compare with other evaluation methods. The results show that the projection pursuit evaluation results of the three kinds of typical cross-sections are 0.571, -0.145 and 0.465 respectively, and the fuzzy comprehensive evaluation results are 0.544, 0.427 and 0.537 respectively. By comparing the above two evaluation methods, it is found that the projection pursuit method can magnify the variability of cross section design factors in the results, thus reducing the impact of subjective factors on the evaluation results and making the selection of urban road cross-section design schemes more scientific and reasonable.

Key words: highway engineering; road cross-section; projection pursuit method; real number coding accelerated genetic algorithm; optimal rejection vector

中图分类号: