基于博弈论的居住区共享停车理论与方法研究

发布时间：2018-01-12 01:13

本文关键词：基于博弈论的居住区共享停车理论与方法研究　出处：《吉林大学》2017年博士论文　论文类型：学位论文

【摘要】：城市机动车保有量剧增与原有停车设施规划不足、配建标准偏低、停车管理技术和方法落后的矛盾导致城市停车困难。因停车拥挤导致车辆巡游时间过长、出入口排队、违规占道停车,进而引发交通纠纷、能源消耗和环境保护等一系列问题。与此同时,因不同类型建筑物停车吸引特性不同导致停车设施利用不均衡的问题同样严重。虽然咪表、VMS等现代化停车装备的引入和停车换乘、阶梯式计价等停车管理措施的实施一定程度可以缓解停车问题,但停车位不足的根源性问题依然无法解决。停车位共享指具有不同属性的相邻土地间共用停车场地,以降低单独使用的总量的一种手段,也是抑制性停车需求控制背景下的产物。但是居住区停车位因其特殊的归属权和管理权问题一直没有被作为共享资源考虑,只有在面临严重的城市停车问题,以及土地资源限制性使用的条件下才开始受到关注。居住区停车位共享策略是利用居住区停车位与其它类型停车位在时间利用方面的互补性来实现停车资源的共享共用,因其庞大的数量和时间方面的特殊优势成为缓解社会停车难的最有利资源。但是居住区停车位又因其特殊的归属权,在共享实施中难免会出现各种各样的问题。本文在前人的研究基础上,将博弈论和共享停车理论相结合,从分析驾驶人、停车位拥有者、停车设施管理者偏好入手,研究居住区停车位参与下的共享停车问题。主要包括以下几方面内容:(1)考虑到传统群体式停车诱导容易导致局部停车拥挤的问题,提出了个性化停车诱导的定义和服务流程,结合居住区共享停车的需求,设计了基于APP的个性化停车诱导系统物理架构和功能框架。在此基础上,研究该框架的几个核心问题。(2)针对居住区停车位能否实现共享这一问题,收集共享停车和博弈理论在交通方面的应用成果,结合国内外共享停车实践,分析论证居住区停车位实施共享的可行性。通过对居住区车辆出行特性的调查分析,研究居住区停车位闲置特性;结合居住区停车位对外共享意愿调研结果,分析居住区停车位实施共享需解决的关键性技术问题、设计规划问题、管理问题和意识问题;在此基础上,建立了居住区停车位对外共享的服务能力评估模型,对模型相关参数进行标定,并提供了可参考数据和模型应用实例,用于评估居住区现有条件下的最大对外共享服务能力,为实时共享停车提供理论基础。(3)由于居住区停车位共享涉及驾驶人、停车位提供者和城市停车设施管理者等多方利益,因停车选择过程中各方的偏好不同,某种程度上会出现利益的分歧,因此引入博弈论对停车选择过程中各方的博弈关系进行研究。文中首先利用共享停车调查结果对影响停车场选择的要素进行分析,从而确立了居住区共享停车中的三方博弈关系,并对博弈模型进行简化处理。分析了几种典型的停车博弈,包括:(1)共享平台和驾驶人之间的动态博弈,给出了停车高峰期二者博弈的基本式,分析了博弈的纳什均衡解;(2)高峰停车中n个驾驶人之间的拥挤停车博弈,通过四个推论分析了由拥挤停车向共享停车转化的基本条件;拥挤停车博弈分析结果表明,拥挤停车产生的根源之一在于违章停车的收益大于共享停车。因此,共享停车的实施务必研究停车秩序管理和违章处罚的阈值。(3)文中通过建立停车秩序管理者和违章驾驶人博弈模型,研究停车秩序管理者和驾驶人博弈的均衡条件。利用案例对几个模型的应用和价值进行说明。针对共享停车中的几种博弈现象,研究了共享停车的博弈均衡:包括n个停车驾驶人的纳什均衡,共同决策下的Cournot均衡,理想状况下的垄断博弈和共享停车动态博弈的Stackelberg均衡。(4)根据博弈分析结果,重点研究个性化共享停车诱导之下的动态博弈行为,利用共享停车选择核心要素构建共享平台和驾驶人之间的Stackelberg博弈双层规划模型,描述二者停车选择博弈行为。重点是基于一定的居住区共享停车假设条件,定义了高峰泊位溢出指数、居住区停车场实际泊位空闲率、居住区停车位高峰对外共享率等评价指标,分别以驾驶人共享停车收益最大化和共享平台协调控制的停车管理效果最佳为目标,建立二者博弈的双层规划模型。通过博弈模型的协调控制,实现二者博弈的均衡,达到区域范围停车场的均衡利用。根据模型的求解条件,设计了基于变形粒子群算法的嵌套优化算法求解博弈模型的均衡解,并以某市商业中心区及周边居住小区为例进行模型仿真分析。本文的创新点包括:(1)构建了居住区停车位对外共享服务能力评估模型,用于描述居住区参与共享停车的服务能力;(2)研究了居住区停车位参与下的共享停车选择行为核心影响要素,分析了驾驶人、停车位供应者和共享停车管理平台之间的博弈关系、博弈模型基本式和博弈均衡解;(3)构建了基于APP共享停车管理框架下的居住区共享停车服务模型,为实现居住区共享停车和停车资源的均衡协调利用提供理论依据。
[Abstract]:City vehicle quantity increase with the original planning of parking facilities, with the construction of low standard, the contradiction between parking management techniques and methods behind difficult to city parking. Due to congestion because of the parking vehicle cruising time is too long, the entrance line, illegal parking, causing a traffic dispute, a series of problems of energy consumption and environmental protection. At the same time. Because of the different types of parking parking facilities with different characteristics to attract the unbalanced problem as serious. Although the meter, VMS and other modern parking equipment and the introduction of park and ride, the implementation of the ladder pricing and parking management measures in a certain extent can alleviate the parking problem, but the shortage of parking spaces of the root causes of the problem still can not solve the parking space sharing. Refers to the adjacent land with different attributes shared parking, a means to reduce the amount of used alone, but also suppression Control of the product under the background of system of parking demand of residential parking spaces. But because of the special ownership and management right has not been considered as a shared resource, only in the face of city parking problem is serious, and the limited land resources use conditions began to attention. Residential area parking sharing strategy the use of residential parking spaces and other types of parking spaces in time by using the complementary aspects of sharing to achieve parking resources, because of the special advantage of its large quantity and time become the most favorable social resources to ease parking difficulties. But the residential parking spaces and because of its special ownership, implementation is in the shared there will be a variety of problems. On the basis of previous research, game theory and shared parking theory, from the analysis of the driver, parking owners, parking facilities management Study on the preference of residential parking spaces in the parking sharing problem. It mainly includes the following aspects: (1) considering the crowded parking problems easily lead to local traditional group parking guidance, put forward the definition of personalized parking guidance and service processes, combined with the residential areas sharing parking demand, design and the function of personalized APP parking guidance system based on the framework of physical architecture. Based on this, several key problems of the framework of the research. (2) for residential parking spaces can realize the sharing, sharing and collecting parking game theory applications in terms of traffic, combined with the practice of parking sharing at home and abroad, analysis of residential area parking space sharing implementation feasibility. Through the investigation on the travel characteristics of residential vehicle analysis of residential area parking idle property; combination of residential parking spaces on the willingness of sharing The results of the survey, the analysis of residential parking implementation of key technical problems needed to be solved sharing, planning, management and awareness of the problem; on this basis, establishes an evaluation model of foreign shared service capacity of residential parking, the relative parameters of the model calibration, and provides reference data and model application for the evaluation of existing residential conditions, the largest foreign share service ability, provide a theoretical basis for the real-time sharing of parking. (3) the residential parking space sharing involves driving, parking providers and city parking facilities management interests, for parking in the process of selecting preferences of different parties, the interests of some the extent of the differences, so the introduction of game theory to study the relationship between the parties in the process of parking choice game. Firstly, using the shared parking survey results on the impact of parking lot. Analysis of optional elements, thus establishing the residential areas sharing three party game relationship in the parking, and the game model is simplified. The analysis of several typical parking game, including: (1) sharing the dynamic game between the platform and the driver, the basic parking peak two game are given, analyzed the game Nash equilibrium solution; (2) the crowded parking game between the N driver in the peak parking, through four deduction analysis by the crowded parking to share the basic conditions of the crowded parking parking transformation; game analysis shows that one of the causes of congestion is the parking parking income is greater than the shared parking. Therefore, you must Study on parking order management and the implementation of shared parking violation penalty threshold. (3) through the establishment of parking order management and illegal driving game model in this paper, study on parking order managers and drivers The equilibrium condition of the game. The application and value of several models by using case. Aiming at sharing several game phenomenon in the parking, the parking sharing game equilibrium: Nash equilibrium includes N parking drivers, Cournot equilibrium joint decision-making under ideal conditions, monopoly game and dynamic game equilibrium parking sharing Stackelberg. (4) according to the results of the game analysis, focus on personalized shared parking guidance under the dynamic game behavior, using shared parking choice key factor in the development of Stackelberg game bi level programming model between the sharing platform and the driver, describe the two parking choice game behavior. The focus is on shared parking residential area based on certain assumptions, the definition of peak berth overflow index, parking lot actual berth idle rate of residential area, residential area parking peak shared rate index, respectively. In order to maximize the revenue sharing drivers parking and sharing platform for coordination and control of parking management is best for the target, bi level planning model is established. Through the coordination of two game control game model, achieve two balance and achieve the balance of the use of area parking lot. According to the solution conditions of the model, is designed based on the deformation of particle swarm algorithm nested optimization algorithm for solving game equilibrium solutions, and a city in central business district and the surrounding residential area as an example for simulation analysis. The innovation of this paper includes: (1) construction of residential parking spaces shared service capacity evaluation model, used to describe the residential area in shared parking service capabilities (; 2) of residential parking spaces in the parking choice behavior of sharing core factors, analyzes the driving, parking and parking management platform providers sharing between Game relations, game models and game equilibrium solutions. (3) a shared parking service model based on APP shared parking management framework is constructed to provide theoretical basis for the balanced and coordinated utilization of shared parking and parking resources in residential areas.

【学位授予单位】：吉林大学
【学位级别】：博士
【学位授予年份】：2017
【分类号】：U491.7

【相似文献】