基于混合演化规则下的零行列式策略演化研究

Evolution of Zero-Determinant Strategies Based on Replication-Aspiration Dynamic

  • 摘要: 在重复囚徒困境中,零行列式策略是一种能够单方面控制博弈双方收益关系的策略,其中剥削策略总是能够获得一个不小于对手的收益。该文基于复制−期望的混合演化规则,研究了合作策略、背叛策略和零行列式策略在方格网络上的演化动力学。通过蒙特卡洛仿真,发现在混合演化规则下,剥削策略能够有效地促进方格网络上合作行为的涌现。通过对演化过程的微观分析,发现“合作−剥削联盟”的存在帮助了合作策略抵制背叛策略的入侵,并且剥削系数对合作策略在方格网络上的扩散有着非单调的作用。

     

    Abstract: In the iterated prisoners’ dilemma game, zero-determinant strategies can unilaterally form a linear relationship between the payoffs of the players, where the extortion strategy always obtains a benefit no less than that of her opponent. We focus on the evolution of the cooperation defection and extortion strategies on the grid network when agents update their strategies by replication-aspiration dynamic. By means of Monte Carlo simulations, we find that the extortion strategy promotes the boost of the cooperation on the grid network under the mixed updating rule. We explain the results by the micro dynamic of the process and find that the existence of "cooperator - extortioner alliance" can help the cooperators resist the invasion of the defectors and the strength of the extortion strategies plays a non-trivial role on the evolution of cooperation.

     

/

返回文章
返回