【作者】 陈娟；

【导师】 张凯； 李继周；

【作者基本信息】 清华大学 ， 物流工程（专业学位）， 2018， 硕士

【摘要】 未来无人驾驶车辆提供出行服务是可预见的交通形式,在人们出行是以购买出行服务的模式之下,购物中心为了吸引更多的消费者,必须保证消费者能够得到及时的出行服务,在此情况下,购物中心停车场需要根据消费者的出行情况,提供合适的无人驾驶车辆,以保障及时满足出行需求尤其是高峰期的出行需求。为此,本文首先全面分析了无人驾驶车辆提供出行服务与现有的出租车、租车提供出行服务之间的差别,在此基础上,假设在无人驾驶车辆提供出行服务之后,人们不会购买私家车,会不会到购物中心消费主要是看该购物中心所属停车场拥有的无人驾驶车辆能否提供及时的出行服务。然后以位于上海郊区——青浦区的桥梓湾购物中心地下停车场为例分析了该停车场在2016年6月的车辆出入停车场数据,全面分析数据得到该购物中心停车场的动态停车特征,构建了停车运行指标体系,再利用排队论方法来建立无人驾驶车辆专用购物中心地下停车场车辆到达率和出口服务率随时间变化的模型,在此基础上结合实际购物中心地下停车场出停车场数据给出该停车场利用排队模型在高峰期时不同服务时间假设前提下在每个单位时间内应该提供的最优车辆数,并利用MATLAB软件画出了在相应的高峰期到达率、服务率和出口数量设置前提下的到达时间和离开时间,等待时间和停留时间图。论文主要工作及研究成果有:(1)分析了上海郊区购物中心地下停车场车辆出入停车场数据,构建了基于该停车场大数据的停车场运行评价指标体系,同时得到该停车场车辆进出停车场规律,对该指标体系反映出的停车场运营问题提出可行的解决方案;(2)将购物中心与无人驾驶车辆两者结合起来,分析现有的购物中心停车场车辆进出停车场的规律,利用排队论方法,建立了无人驾驶车辆专用的购物中心地下停车场的车辆到达率和出口服务率随时间变化的模型;(3)基于实际购物中心地下停车场出停车场数据,利用排队模型,研究了高峰期的单位时间内应提供的最优车辆数,以及与高峰期到达率、服务率的关系。更多还原

【Abstract】 Providing travel services for unmanned vehicles in the future is a foreseeable form of transportation.Under the mode that people travel by buying travel services,the mall in order to attract more consumers,consumers must be guaranteed timely travel services.Under the circumstances,the shopping mall parking lot needs to provide suitable unmanned vehicles according to the consumers’ travel conditions,so as to guarantee the timely satisfaction of travel demand,especially during the peak period.Therefore,this article first comprehensively analyzes the differences between the provision of travel services by unmanned vehicles and existing taxis and car rental services.Based on this,it is assumed that after the driverless vehicles provide travel services,people will not When buying a private car,whether it will go to the shopping mall mainly depends on whether the self-driving vehicles owned by the mall’s owned parking lot can provide timely travel services.Then,taking the underground parking lot of Qiaobu Bay Shopping Center in the suburbs of Shanghai,Qingpu District,as an example,the vehicle parking data of the parking lot in June 2016 was analyzed.The comprehensive analysis of the data obtained the dynamic parking feature of the shopping mall parking lot.A parking operation index system was constructed,and a queuing theory method was used to establish a model of the vehicle arrival rate and export service rate of an underground parking lot for a shopping center dedicated to unmanned vehicles over time.Based on this model,the actual underground parking lot of the shopping center was integrated.The parking lot data shows the optimal number of vehicles that should be provided in each unit time during the peak period using the queuing model for different service time assumptions,and draws the corresponding peak arrival rate using MATLAB software.Arrival time and departure time,waiting time and dwell time chart under the premise of setting of service rate and export quantity.The main work and research results of the paper are:(1)Analyzed the data of the parking lot entering and leaving the parking lot of the underground parking lot in the suburbs of Shanghai,built the evaluation index system of the parking lot operation based on the parking lot big data,and obtained the regularityof the parking lot vehicle entering and leaving the parking lot,reflecting the index system Out of the problem of parking operations to propose feasible solutions;(2)Combine the shopping center with the driverless vehicle to analyze the existing rules for entering and leaving the parking lot in the parking lot of the shopping mall,and use the queuing theory to establish an underground parking lot for shopping centers dedicated to driverless vehicles.Models of vehicle arrival rates and export service rates over time;(3)Using the queuing model,based on the data of the underground parking lot out of the actual shopping center,the queuing model was used to study the relationship between the optimal number of vehicles that should be provided during the peak period and the arrival rate and service rate in the peak period.更多还原