搜索

x
中国物理学会期刊

容忍驱动迁移机制下人机混合三策略空间囚徒困境中的合作演化

Evolution of cooperation in a human-machine hybrid three-strategy spatial prisoner’s dilemma game with tolerance-driven migration

PDF
导出引用
  • 本文在二维格子网络上构建了一个引入容忍驱动迁移机制的人机混合三策略空间囚徒困境博弈模型.该模型中包含合作机器人与普通个体两类主体:合作机器人在整个演化过程中始终采取合作行为,而普通个体可在合作、背叛或独行三种策略中进行选择.通过引入容忍驱动迁移机制,系统探讨了容忍度、机器人比例、群体密度和困境强度对合作演化的影响.研究结果表明,存在一个最优群体密度能使群体合作水平达到峰值.适度的空位为个体提供了必要的迁移条件,而较低的容忍度则增强了个体对背叛环境的敏感性,二者协同促进了合作团簇的形成与维持.相反,群体密度过高将导致个体迁移受阻,使其难以有效逃离背叛环境,进而引发合作崩溃.此外,我们发现合作机器人的作用与群体密度和困境强度密切相关:在困境强度和群体密度较高时,引入一定比例的合作机器人反而会抑制群体合作行为.进一步地,本文还探究了具有更大策略空间的社交机器人以及不同博弈类型对群体合作演化的影响,同样发现存在一个最优群体密度使合作水平最高.上述发现为我们理解人机混合系统中的合作演化规律提供了新的视角.

     

    We propose a human-machine hybrid three-strategy spatial prisoner’s dilemma game model on a two-dimensional lattice with a tolerance-driven migration mechanism. The population consists of cooperative robots and ordinary individuals, where robots remain cooperative throughout evolution, while ordinary individuals can adopt cooperation, defection, or the loner strategy. We then investigate how the tolerance threshold, robot proportion, population density, and dilemma strength affect the evolution of cooperation. Results show that cooperation is maximized at an optimal population density. Moderate vacancy provides suffcient space for migration and spatial reorganization, and a lower tolerance threshold increases individuals’ responsiveness to defector-dominated neighborhoods; together, these effects facilitate the formation and stability of cooperative clusters. By contrast, excessive population density hinders migration, making it diffcult for individuals to escape defector-dominated neighborhoods and thus causing cooperation to break down. We further show that the effect of cooperative robots depends on both population density and dilemma strength. The role of cooperative robots is not always positive, under high dilemma strength and high population density, the introduction of cooperative robots can even suppress cooperation. Furthermore, we examine the effects of robots with a broader strategy space—specifically, robots that adaptively switch between cooperation and the loner strategy under high local defection pressure—and different game types on the evolution of cooperation, and similarly find that an optimal population density maximizes cooperation. The adaptive strategy switching reduces defector exploitation of robots and sustains higher cooperation levels under strong dilemmas. Robustness checks under the snowdrift game and the weak prisoner’s dilemma confirm that the unimodal density–cooperation relationship is a general feature independent of the specific game type. These findings deepen our understanding of the evolution of cooperation in human-machine hybrid systems.

     

    目录

    /

    返回文章
    返回