电动汽车分时租赁车辆调度和推荐策略的研究与设计

发布时间：2018-06-10 04:42

本文选题：车联网 + 电动汽车　；参考：《北京邮电大学》2016年硕士论文

【摘要】：电动汽车作为绿色环保的新能源交通工具已受到各国的广泛关注,随着分时租赁新型商业模式的不断推广,电动汽车正在慢慢渗透进民众的生活,成为大众绿色出行的新选择。分时租赁运营模式下的车辆使用时长短、高峰时段集中、支持异地还车等特点显著,经典的车辆租赁调度算法便不再很好地适用。目前已有的电动汽车调度算法也较为简略和笼统,典型算法为受限调度与控制调度两类,受限调度算法不支持跨站调度;控制调度算法单一地将运营收益作为车辆调度和分配策略的主要指标。这两种调度算法在分时租赁场景下都具有相当的局限性,因此电动汽车分时租赁调度策略算法的优化和完善存在很大空间。本论文以电动汽车运营服务支撑环境和分时租赁平台为依托,针对已有调度算法的局限性,从服务率、车辆日均用车时间分布、运营收益和平均调度距离等各方面进行优化与改进,提出了基于多目标最优化模型的“全服务”调度模式,并以北京绿狗租车分时租赁系统的站点网络和车辆信息为仿真环境和数据集,仿真验证和分析了算法的有效性。同时,本文还系统的介绍了“全服务”车辆调度与推荐分配策略下的电动汽车分时租赁平台的设计与实现。论文首先概要性地介绍了电动汽车分时租赁的商业模式,之后详细介绍了本文的创新性算法——“全服务”车辆调度和推荐策略,从算法的需求分析、问题描述、数学建模、仿真与结果分析等方面进行了详尽描述。接下来,按照软件工程的流程,对基于“全服务”车辆调度和推荐策略的分时租赁平台进行了细致的需求分析、系统总体设计、详细设计与实现、测试与验证等。论文结尾对全文工作进行了梳理总结,并提出了下一步的研究方向。
[Abstract]:As a new energy transportation tool for green and environmental protection, electric vehicles have attracted wide attention from all countries. With the continuous promotion of new type of time leasing business model, electric vehicles are slowly infiltrating into the lives of the people, becoming a new choice for the mass green travel. The classic vehicle rental scheduling algorithm is no longer well applicable. The existing scheduling algorithms for electric vehicles are also relatively simple and general. The typical algorithms are limited scheduling and control scheduling, and the limited scheduling algorithm does not support the cross station scheduling; the control scheduling algorithm uses the operating income as a vehicle to tune a single vehicle. The main index of degree and allocation strategy. The two scheduling algorithms have considerable limitations in the time sharing rental scenario. Therefore, there is a lot of space for the optimization and improvement of the algorithm for the time sharing and scheduling of electric vehicles. This paper is based on the support environment of the electric vehicle operation service and the time sharing lease platform. Limitations are optimized and improved from service rate, vehicle daily vehicle time distribution, operating income and average scheduling distance. A "full service" scheduling model based on multi-objective optimization model is proposed, and the site network and vehicle information of the Beijing green dog renting time rental system are simulated as the simulation environment and data set, and the simulation is simulated. The effectiveness of the algorithm is verified and analyzed. At the same time, the design and implementation of the "full service" vehicle scheduling and the recommendation allocation strategy for the time sharing rental platform for electric vehicles is introduced in this paper. First, the paper briefly introduces the business model of the time sharing lease for electric vehicles, and then introduces the innovative algorithm in detail. "Full service" vehicle scheduling and recommendation strategies are described in detail from the requirements analysis, problem description, mathematical modeling, simulation and result analysis in the algorithm. Next, detailed requirements analysis is made for the time sharing rental platform based on the "full service" vehicle scheduling and recommendation strategy. Design, detailed design and implementation, testing and verification. At the end of the paper, the whole work is summarized, and the next research direction is put forward.
【学位授予单位】：北京邮电大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP391.3

【参考文献】