Cooperation Emergence Mechanism Research Under Network Evolutionary Public Goods Game

The classical Nash equilibrium theory predicts the result of the game based on the complete rationality of the individuals participating in the game.With the introduction of game theory into the study of evolutionary biology,individual behaviors and population thinking patterns are also incorporated into game theory;evolutionary game theory no longer relies on the assumption of rational individuals;the more fitness strategies in evolution,the greater the chance of that imitated and learned.With the development of complex networks,network evolutionary games is the frontier research field of game theory.The public goods game is a typical social dilemma game,from the perspective of social dilemmas,as a kind of extention from the dilemma game of two prisoners to multiplayer games,the public goods game depicts the contradiction between self-interest and collective cooperation,in the study of the cooperative emergence of evolutionary game,the game has attracted wide attention.In view of the characteristics of the public goods game,this paper presents the mechanism of group learning,the research focuses on: in the evolution of network games,the influence of different strategy updating methods on the emergence or inhibition of group cooperation in evolutionary games.According to the learning mechanisms proposed in this paper and the comparative analysis of the existing classical learning mechanisms,the results obtained are as follows:(1)It is pointed out that group learning mechanism is the characteristic of public goods game,and the classical local learning mechanism is more suitable for the prisoner's dilemma;(2)Under the group learning rules,the strategy of learning the individual with the maximum payoff in a neighborhood can make the cooperation level being higher than the local learning rules on square lattices,nearest-neighbor coupled networks,and WS small-word networks;(3)Under the group learning rules,the strategy of learning the random individual in a neighborhood can make the cooperation level being higher than the local learning rules on nearest-neighbor coupled networks,and WS small-word networks;(4)Compare the use of penalty,reward mechanisms alone and reward cooperators while punishing defectors,simulation results show that punishing defectors while rewarding cooperators mechanism is superior to the individual using punishment or reward mechanism under the same parameters;combined with the reward and punishment mechanism,in most cases,the group learning rules are more excellent than the local learning rules;With small punishments,one can achieve higher levels of group cooperation than that of rewards.