大型展览活动人群智能服务管理研究

【作者】 李进；

【导师】 朱道立；

【作者基本信息】 复旦大学 ， 管理科学与工程， 2010， 博士

【摘要】 近年,随着我国经济文化事业的快速发展和综合国力的进一步增强,我国举办的各类大型展览活动越来越多,例如广交会、世博会等,这些活动日益成为社会、企业和广大民众关注的焦点。大型展览活动参与人数多、社会反响大,一般具有人群密集性、人群不可预测性和人群分布不均匀性等特征,安全和服务问题突出。切实加强大型展览活动人群服务管理能力、预防和处理人群事故是达到高性能的安全保障和高质量的服务效果的重要内容。以2010年上海世博会为契机,我国开始逐步加强大型展览活动人群服务管理体系的建设,中央和地方都根据各自的特征建立了相应的大型展览活动管理组织机构,全国性的人群服务管理体系框架已初步形成,并且在人群事故预防和人群服务实践中已经发挥着很重要的作用。然而,目前的人群服务管理体系的建设主要是从宏观层面进行的,它只能表明大型展览活动人群服务管理已经引起政府相关层面的重视。面对大型展览活动人群问题的突发性和不确定性,我国现有的人群服务管理体系在人群服务、人群事故的处置和管理上还存在着许多问题,难以有效地应对各种人群问题。必须从解决微观层面的问题尤其是难题入手,对相关的理论、技术和方法进行创新性地研究,给出最优的解决方案,从而全面提高我国大型展览活动人群服务管理水平。本文综合通信和计算机科学、管理科学、运筹学等学科,集成移动通信、情景感知、系统仿真等技术,系统地研究大型展览活动人群智能服务管理问题,揭示人群服务管理中微观和宏观层面之间的内在联系,探索大型展览活动人群智能服务管理规律和机理。根据游客参观全过程的服务需求分析,从参观计划、参观协调、服务保障和应急反应的决策问题着手,研究了大型展览活动人群智能服务管理中的微观层面的关键问题。各章主要内容如下：第一章绪论,介绍了本文的研究背景、研究内容、研究意义及论文结构。第二章文献综述,概括了大型展览活动人群智能服务管理相关概念,从情景感知系统在游览活动中的应用、人群运动和大型展览活动人群服务运营三个方面对大型展览活动人群智能服务管理的相关研究进行了综述,并在此基础上,进行了简要的评述,指出了目前大型展览活动人群服务管理的研究不足。第三章,从系统论的角度研究了大型展览活动人群智能服务管理问题。首先,提出了大型展览活动人群智能服务系统的定义,介绍了系统的基本维度,包括系统的要素、系统的目标、系统的结构、系统的状态和系统的环境,阐述了系统信息感知与发布技术。然后,提出需要深入研究的系统控制与优化问题,并且建立了大型展览活动服务准备、服务实施和应急控制三位一体面向全过程的综合服务运营管理体系。最后,在人群服务需求分析的基础上设计了大型展览活动人群智能服务决策支持系统,概述了系统的整体结构,介绍了系统的主要功能,并提出了系统数据库的设计构想。第四章,从游客参观行程计划的角度研究了时问依赖网络中游客个性化参观行程设计问题。考虑到人群拥挤、突发事件和道路类型对出行时间的影响,提出了基于离散时间的时间序列图模型,通过时间序列图模型表示交通网络中不同时间不同路段的出行时间。在此基础上,提出了两种既符合大型展览活动交通实际,又符合游客个人偏好的参观行程设计方案,即满足游客单日和多日参观计划需求约束下的最优行程设计,分别建立了相应的混合整数规划模型,设计了基于网络计划和动态规划思想的标号算法。第五章,从游客参观协调的角度研究了基于Multi-agent的人群服务协调控制问题。首先,设计了基于Multi-agent的人群智能服务系统,概述了系统的基本结构,详细介绍了Multi-agent系统的建模思路。然后,提出了多阶段动态协调控制方法,该方法在引入帕累托最优概念的基础上,通过游客总体满意度下的择优选择机制为游客实时动态的决定下一参观目标点的选择,使用Logit模型进行参观路线的选择。最后,以上海世博会为背景,设计了基于Multi-agent系统的仿真试验和三种对比方案,对协调控制方法进行了检验和灵敏度分析。第六章,从服务保障的角度研究了大型展览活动移动资源调度问题。介绍了移动资源调度工作的概念模型,总结了移动资源调度的过程。在此基础上,针对大型展览活动中服务资源需求的多样性和预测的资源需求情形,研究了考虑未来需求的多资源调度问题,建立了基于期望成本的多资源调度模型,基于图论中网络优化和线性规划优化思想,设计了多资源分类调度的启发式算法,并分析了算法复杂性。第七章,从突发事件应急反应的角度研究了大型展览活动人群疏散问题。介绍了大型展览活动中的突发事件,分析了突发事件的特征,对人群事故发生机理进行了初步的探索。考虑到大型展览活动突发事件应急疏散时,常常出现多源点同时需要疏散,且有多个疏散汇点的情形,研究了多源多汇和有容量限制条件下的应急疏散问题,建立了基于路径的网络流控制疏散模型,设计了基于关键优选路径进行网络流控制的启发式算法,并分析了算法的复杂性。最后,在结论部分,对论文的工作进行了总结,并对未来的研究加以展望。更多还原

【Abstract】 In recent years, as the rapid development of China’s economic and cultural undertakings and further enhancement of China’s overall national strength, various large-scale exhibitons organized by China are increasing, e.g. Chinese Export Commodities Fair and World Expo. These exhibitons are increasingly becoming the focus of attention paid by society, businesses and the general public. Large-scale exhibitons have a large number of participants and great social repercussions. The crowd’s characteristics of the large-scale exhibitons are intensive, unpredictable and uneven distribution. So, security and service issues are very obvious. It becomes an urgent task to enhance the ability of crowd’s service management, prevent and deal with crowd incidents so as to achieve high-performance security and high-quality service.Taking 2010 Shanghai World Expo as an opportunity, China has gradually built the crowd service management system for large-scale exhitions with the establishment of respective organizations in charge of crowd service management in national,provice and city levels.The national crowd service management system of large-scale exhibitons has begun to take shape and play an important role in practice. However, since the current crowd service management system is designed at macro level, it is far from the practical need at the micro level. The theory, technologies and methods at the micro level are the base of the whole crowd service management system. We should focus on the research in micro-level problems especially difficult problems. Further study on revelent theory, technologies and methods will surely provide an optimal solution so as to improve the overall crowd service management level of Chinese large-scle exhibitons.This dissertation employs multiple research methods including communication and computer science, management science and operations research, etc. and integrates a lot of technologies, such as mobile communicaition, context aware and system simulation. It is aimed to systematically study the crowd intelligent service management issues of large-scale exhibitons, unveil the inner correlation between the micro and macro levels, and finally provide new perspective and decision support for crowd intelligent service management of large-scale exhibitions in the future. According to analysis of tourist service demand in the whole trip, based on the decision-making problems in the process of crowd service management including travel plan, travel coordination, service security and emergency response, this dissertation studys the micro-level key problems involved in crowd intelligent service management of large-scale exhitions. The main contents of this dissertation are as follows:In chapter 1,the research background, contents and significance have been introduced together with the structure of the dissertation. In chapter 2,literature review is made to summarize the concepts related to crowd intelligent service management in large-scale exhibitions, and cover the research conducted from three aspects:applications of context-aware systems in travel activities, crowd movement and crowd service operations in large-scale exhibitions. Furthermore, we commentate on the previous research and point out the deficiency in the current research on crowd service management in large-scale exhibitions.In chapter 3,this dissertation studies the crowd intelligent service management in large-scale exhibitions from the perspective of system theory. Firstly, we propose the definition of crowd intelligent service system for large-scale exhibitions, and introduce the system’s basic dimensions including system elements, objectives, structure, status and environment. System information awareness and release technologies are also introduced. Then, this dissertation gives the system control and optimization problems deserved to be studied deeply with the establishment of the whole process-oriented integrated service operation management system including three levels of large-scale exhibitions:service reserve, service implementation and emergency control.Finally, based on the analysis of crowd service needs, crowd intelligent service decision-making system for large-scale exhibitons is designed. The whole systematic organization is illustrated. The main functions and a blueprint for system database design are introduced.In chapter 4, this dissertation mainly focuses on personalized tourist trip design based on time-dependent network from the perspective of tourist trip plan. Taking into account the impact of such factors as crowd congestion, emergency, and road types on travel time, we propose a time-series graph model based on discrete time. The travel time in transport network in different sections and different time is represented using the time-series graph model. On this basis, we propose two trip design methods which are consistent with the current traffic and tourist personal preference in large-scale exhibitions, that is, one-day tourist trip design method and multi-day tourist trip design method. Then, we respectively establish the mixed integer programming models and design labeling algorithms based on the idea of network planning and dynamic programming. In chapter 5, crowd service coordination and control based on multi-agent is considered from the perspective of travel coordination. Firstly, crowd intelligent service system based on multi-agent is designed, the whole systematic organization is illustrated, and the ideas of modeling the multi-agent system are introduced in detail. Moreover, we propose a multi-stage dynamic coordination and control method. Through introducing the concept of Pareto Optimality, this method decides the next visiting node in real time and dynamically for tourists based on preferential selection mechanism to total tourist satisfaction value, and selects roads using Logit model. Finally, under the background of Shanghai World Expo, multi-agent system simulation and three different methods are designed, the coordination and control methods are tested, and sensitivity analysis is also making.In chapter 6, mobile resource scheduling problem of large-scale exhibition is investigated from the perspective of service security. We introduce a conceptual model of mobile resource scheduling work, and give the process of mobile resource scheduling. On this basis, due to the diversity of service resources and forecast of resource demands, this chapter studies the multi-resource scheduling problem considering future demand, and establishes a multi-resource scheduling model based on expectation cost. Furthermore, a novel heuristic algorithm based on network optimization in graph theory and linear programming optimization is given in terms of multi-resource classification. The computational complexity of the algorithm is also analyzed.In chapter 7, crowd evacuation of large-scale exhibitions is considered from the perspective of emergency response. We introduce the emergent incidents in large-sclae exhibitions, analyze the characteristics of the incidents, and start a preliminary study to mechanism of crowd incidents. Then, we study the emergency multi-source and multi-destination evacuation problem with the capacity constraints, and propose a path-based network flow control evacuation model. A novel heuristic algorithm of network flow control is designed based on key selected path. The computational complexity of the algorithm is also analyzed.In the conclusion, we pointed out the main innovation of the dissertation and the perspective for the further research.更多还原