基于囚徒困境策略的改进HK网络上的合作博弈

摘要

为模拟现实世界的合作行为,本文在HK网络模型基础上提出了一种具有高聚类幂律可调性质的新的网络模型,并分析了囚徒困境博弈在此网络上的演化.通过仿真实验,研究了该网络的高聚类特性对合作行为的影响.大量实验表明,网络的高聚类特性可以极大促进合作现象的涌现.同时研究也发现,随着诱惑参数的变大,合作水平也会随之下降,但幅度不大.总之,该演化博弈模型可以促进合作现象的涌现并抵御背叛策略的传播.%To simulate the cooperative behavior using the HK network model,we proposed a new adjustable power-law high-clustering network model to analyze the prisoner dilemma game.Through simulations,we investigated the effect of high clustering on the cooperative behavior.The experimental findings suggested that high clustering coefficient values may considerably contribute toward the emergence of cooperative behavior.We also found that with increasing temptation,the level of cooperation decreased and the variation was small.Altogether,the evolutionary game model promotes cooperation and hinders betrayal.