地球信息科学学报 ›› 2021, Vol. 23 ›› Issue (5): 802-811.doi: 10.12082/dqxxkx.2021.200360

• 地球信息科学理论与方法 • 上一篇    下一篇

电动出租车专用充电场站选址模型研究

张毅1,*(), 朱攀2   

  1. 1.信息工程大学,郑州 450001
    2.加华地学(武汉)数字技术有限公司,武汉 430223
  • 收稿日期:2020-07-09 修回日期:2020-09-17 出版日期:2021-05-25 发布日期:2021-07-25
  • 通讯作者: 张毅
  • 作者简介:张 毅(1976— ),男,甘肃白银人,博士,副教授,主要从事GIS时空建模与分析应用研究。E-mail:jkkl126@126.com
  • 基金资助:
    国家重点研发计划项目(2016YFB0502300)

Research on Site Selection Model of Special Charging Stations for Taxis

ZHANG Yi1,*(), ZHU Pan2   

  1. 1. Information Engineering University, Zhengzhou 450001, China
    2. Can-Cn Geo-digitization Technology Company Limited, Wuhan 430223, China
  • Received:2020-07-09 Revised:2020-09-17 Online:2021-05-25 Published:2021-07-25
  • Contact: ZHANG Yi
  • Supported by:
    National Key R&D Program of China(2016YFB0502300)

摘要:

近年来随着能源短缺和环境污染问题日益严重,低排放、低噪声、节能环保的电动汽车受到了越来越多的关注。特别在电动出租车领域,我国公共充电设施的规划和建设已迫在眉睫。针对目前电动出租车专用充电站选址研究方面存在的缺少丰富详实的车辆运行数据支撑、未能充分结合出租车的换班特性开展分析和在选址模型构建方面未能充分考虑已建成充电站的影响等不足,本文提出了一种合理可行的电动出租车专用充电场站选址方案,具体流程为:① 基于对出租车GPS数据的统计和分析,利用蒙特卡洛方法提取充电需求;② 对充电站建设的影响因素进行分析,综合考虑了建设成本,出租车等待成本和已建成充电站的影响,建立了充电站的选址模型;③ 利用多种群遗传算法对模型进行求解;④ 利用北京市1.2万辆出租车连续一周的运行轨迹数据提取了充电需求,模拟10万辆电动出租车的充电需求规模对选址模型进行了算例分析。实验结果表明,选址模型能够有效地减少充电站建设投入的成本并缩短出租车平均充电等待时间。

关键词: 充电需求, 电动出租车, 出行链, 充电站, 选址, 建设成本, 等待时间, 多种群遗传算法

Abstract:

With the increasing shortage of energy and occurrence of various environmental pollution problems in recent years, low-emission, low-noise, energy-saving, and environmentally friendly electric vehicles have received more and more attention nowadays. In the field of electric taxis in China, the planning and construction of public charging facilities are imminent. At present, there are still deficiencies in the research on site selection of electric taxi charging stations, such as the lack of rich and detailed trip data, insufficient analysis based on trip characteristics and operation state of taxis, and insufficient consideration of the impact of existing charging stations. Thus, this paper proposes a reasonable and feasible site selection framework for taxi charging stations. Based on the statistical analysis of GPS data of taxis, the Monte Carlo method is used to derive charging demand. The influencing factors of charging station construction including the construction cost, charging time, and the impact of established charging stations are analyzed to establish the site selection model of charging stations. And this model is further solved using Multi-Population Genetic Algorithm (MPGA). We obtained the trajectory data of 12 000 taxis in Beijing in one week to derive the charging demand. The site selection model was further established based on the charging demand of 100000 electric taxis. Our results show that the site selection model was feasible, which can effectively reduce the cost of charging station construction and shorten the average wait time of taxi charging.

Key words: charging demand, electric taxis, trip chain, charging stations, site selection, construction cost, waiting time, Multi-Population Genetic Algorithm