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）

重庆交通大学学报（自然科学版） ›› 2025, Vol. 44 ›› Issue (11): 44-52.DOI: 10.3969/j.issn.1674-0696.2025.11.06

• 交通运输+人工智能 • 上一篇    

城市出租车寻客热点区域挖掘与优选推荐方法

周旦1，娄本潇1，顾国斌2，钟楚捷3，孙家煜4   

  1. (1. 桂林电子科技大学 建筑与交通工程学院，广西 桂林 541004； 2.南宁学院 广西中国-东盟综合交通国际联合重点实验室， 广西 南宁 530200； 3. 广州市交通规划研究院有限公司，广东 深圳 510180； 4. 宁波大学 海运学院，浙江 宁波 315211)
  • 收稿日期:2023-11-13 修回日期:2025-05-08 发布日期:2025-11-27
  • 作者简介:周旦(1978—)，男，江西修水人，副教授，博士，主要从事交通大数据挖掘与分析方面的研究。E-mail：8697992@qq.com 通信作者：顾国斌（1997—），男，江苏南通人，博士研究生，主要从事轨迹数据挖掘和无人机计算机视觉方面的研究。E-mail:15251707205@163.com
  • 基金资助:
    国家自然科学基金项目(71861005)；南宁市科学研究与技术开发计划项目（20223230）；广西自然科学基金项目（2020GXNSFAA159153）；广西高校中青年教师科研基础能力提升项目（2024KY1883，2024KY1180）

Mining of Hotspot Areas for Urban Taxi Passengers-Seeking and Recommendation Method for Preferential Selection

ZHOU Dan1, LOU Benxiao1, GU Guobin2, ZHONG Chujie3, SUN Jiayu4   

  1. (1. School of Architecture and Transportation Engineering, Guilin University of Electronic Technology, Guilin 541004, Guangxi,China; 2. Guangxi Key Laboratory of International Join for China-ASEAN Comprehensive Transportation, Nanning University, Nanning 530200, Guangxi, China; 3.Guangzhou Transport Planning Research Institute Co., Ltd., Shenzhen 510180, Guangdong, China; 4. Maritime College, Ningbo University, Ningbo 315211, Zhejiang, China)
  • Received:2023-11-13 Revised:2025-05-08 Published:2025-11-27

PDF

摘要： 为解决出租车寻客区域评价指标单一等问题，提出引入毕达哥拉斯模糊诱导有序加权对数平均距离算子的综合评价模型。该方法借助基于网格密度的GSCAN空间聚类算法挖掘海量轨迹数据中蕴含的上下客热区信息，进而提出一种由运营指标和道路指标共同构成的指标体系完成对寻客区域的评价。采用成都市某一工作日早高峰时段的出租车轨迹数据，挖掘成都市工作日早高峰时段高载客概率区域分布情况以及评价优选推荐区域。结果表明：以东通社区西门为寻客起点的最优目的地是成都东站区域，该区域载客率达57.8%，较备选区域平均水平高28.0%，每公里红绿灯数量较平均水平低9.0%，与成都东站出租车需求量大、道路便乘设施完善的实际相匹配。研究成果可为出租车司机寻客提供决策依据。

关键词: 交通运输工程；出租车寻客；热点识别；寻客成功概率指标体系；模糊评价

Abstract: In order to solve the problem of single evaluation index for taxi passenger-seeking area, a comprehensive evaluation model introducing Pythagorean fuzzy induced ordered weighted logarithmic mean distance operator was proposed. With the help of GSCAN spatial clustering algorithm based on grid density, the proposed method mined the hotspot information of passenger boarding and alighting contained in the massive trajectory data, and then an index system consisting of operation indexes and road indexes was proposed to complete the evaluation of passenger-seeking area. The taxi trajectory data from a weekday morning rush hour in Chengdu, the distribution of areas with high probability of carrying passengers during weekday morning peak hour in Chengdu was mined and the preferred recommended areas were evaluated. The results show that the optimal destination with the west gate of the Dongtong community as the starting point of passenger-seeking is the Chengdu East Railway Station area, which has a passenger carrying rate of 57.8%, which is 28.0% higher than the average level of the alternative areas, and the number of traffic lights per kilometer is 9.0% lower than the average level, which matched the actual situation of Chengdu East Railway Station, where there is a high demand for taxis and well-equipped road facilities. The research results can provide a decision-making basis for the taxi drivers to search for passengers.

Key words: traffic and transportation engineering; taxi passenger-seeking; hotspot identification; success probability indicator system of passenger-seeking; fuzzy evaluation

中图分类号: