【作者】 陈伟；

【导师】 王建伟；

【作者基本信息】 东北大学 ， 管理科学与工程， 2020， 硕士

【摘要】 无论是古代社会还是现代社会,人类一直受到各种各样的传染病威胁,每次流行病大规模爆发时,都会严重影响人们的生命财产安全,给社会带来巨大的管理负担,严重地阻碍社会的发展进步。随着科学技术的发展,医学工作者研制了各种应对不同流行病毒的疫苗,以在从根本上阻止流行病的蔓延。然而,接种疫苗需要耗费一定的成本,采取自愿的方式会使得人们心存侥幸从而导致接种意愿较低,最后并不能有效阻止流行病的大规模蔓延。因此,研究社会中影响个体的接种意愿的因素和流行病传播机制,进而采取有效措施降低社会染病规模是极其重要的。以往对于接种行为的研究主要借助于经典理论框架,近几年来,演化博弈理论和复杂网络理论的提出为接种博弈的研究提供了新的视角。同时,以往对于社会网络接种博弈的研究主要集中对个体单一博弈行为的研究,很少涉及个体不同行为之间的相互影响。因此,本文针对于现有背景以及已有研究成果存在的局限性,从以下几个方面进行了相关的研究。(1)首先,研究了双层社会网络上信息共享驱动接种博弈演化的过程。本文从现实背景出发,考虑个体的接种策略的选择不仅与自身的收益有关,还可能会受到其社交网络上自己的亲朋好友的影响,即由于信息共享的影响使得个体策略学习范围扩大,最后影响个体的策略选择。据此构建了双层网络上信息共享驱动接种博弈演化模型,同时通过计算机仿真发现信息共享可以在一定程度上减小网络中的染病规模。(2)其次,研究了双层社会网络上信息共享与接种博弈协同演化过程。本文在信息共享驱动接种博弈演化模型的基础上,考虑了接种行为对信息共享的影响,即个体收益的不同使得个体的信息共享变得有选择性。据此建立了双层网络上信息共享与接种博弈协同演化模型。通过计算机仿真分析发现,当共享程度较小时,信息共享可以减小网络中的染病规模;当信息共享较大时,信息共享反而使得网络个体染病比例更大。(3)最后,研究了考虑信息共享的接种博弈演化模型在PGP信任网络结构上演化过程。本文选取现实存在的PGP信任网络结构对以上接种博弈演化模型进行数值仿真,并与前面章节的模型结果进行对比分析。本文发现,真实社会网络的聚类效应会使得信息共享的单向作用对染病规模的抑制起到负面作用;而当信息共享具有异质性时信息共享对于染病规模的抑制会起到积极作用。本文提出的考虑信息共享的双层社会网络上接种博弈模型,不仅在模型背景上更贴合实际,反映个体接种博弈行为的演化过程,而且通过利用真实社会网络结构对所构建的模型进行仿真模拟,对比分析了造成结果差异的原因。为相关医疗卫生部门提供更精准的管理工具,也为后续的研究工作提供新的视角。更多还原

【Abstract】 Human society has always been threatened by various infectious diseases both in ancient times and modern times.It will demage people’s lives and property seriously,bring huge administrative burden to the society and hinder Social development seriously when an epidemic breaking out.With the development of science and technology,medical workers have developed various vaccines to deal with different epidemic viruses to fundamentally prevent the spread of epidemics.Because of requiring a certain cost,vaccinating freely leads people not to vaccinate with a fluke mind which can not prevent the spread of epidemics.It is extremely important to study the factors affecting the vaccination willingness of individuals and the mechanism of epidemics in society which we can provide effective measures to reduce the scale of social illness in the process of studying.The past research on vaccination behavior mainly relied on the classical theoretical framework until the evolutionary game theory and the theory of complex network theory have been introduced which provid a new perspective for the research of vaccination game.At the same time,the previous research on social network vaccination game mainly focused on the study of individual single game behavior,and rarely involved the interaction between individual behaviors.Therefore,in view of the existing background and the limitations of existing research results,this paper has carried out related research from the following aspects:Firstly,we sutdy the evolution of information sharing-driven vaccination games on a twolayer network.Based on realistic background,we consider that the choice of an individual’s vaccination strategy is not only related to its own benefits,but also be affected by its relatives and friends on its social network,that is,the scope of individual strategy learning is expanded due to the impact of information sharing which affects individual’s strategy choices.And then we constructe an evolution model of the information sharing-driven vaccination game on the double-layer network.At the same time,through computer simulation,we find that information sharing can reduce the scale of disease in the network to a certain extent.(2)Secondly,we studied the co-evolution process of information sharing and inoculation games on a two-layer network.Based on the evolutionary model of information sharing-driven vaccination games,we consider the impact of vaccination behavior on information sharing--the difference in individual returns makes individual information sharing selective.And then wen contructed a co-evolution model of information sharing and vaccination games on a two-layer network.Through computer simulation analysis,we found that when the probability of information sharing is small can reduce the scale of disease infection in the network,and the result is opposite when the information sharing is large.(3)Finally,we studied the evolution process of the inoculation game evolution model considering information sharing on the structure of the PGP trust network.We select the existing PGP trust network structure to simulate the evolution model of the above inoculation game,and compares it with the model results in the previous chapter.We find that the one-way effect of information sharing negatively affect the scale of infection with clustering.And information sharing will have a positive effect on the scale of infection when information sharing is heterogeneous.The vaccination game model on the two-layer network considering information sharing proposed in this paper not only is realistic,but also reflects the evolution process of individual vaccination game behavior realistically.We use the real social network structure to build the model and find the reasons for the difference in results through the comparative analysis so that we can provide more accurate management tools for relevant health care departments and a new perspective for subsequent research work.更多还原