Evolutionary Mechanism Of Group Cooperation Based On Reputation Classification

The theory of "survival of the fittest" indicates that competition in resource-limited environments is the main driver of species' evolution.However,cooperative behaviour widely exists in the nature and society.How to understand the contradictory relationship between the competition among selfish individuals and the cooperation that exists widely in real life has attracted the close attention of academic.For many years,scholars have devoted themselves to exploring the causes of cooperation and the methods to promote cooperation.Among which evolutionary game theory provides a powerful theoretical framework to explain the existence of cooperative behavior.Nowak,a biologist from Harvard University,summarized five typical mechanisms for promoting cooperation,which are kin selection,direct reciprocity,indirect reciprocity,group selection and network reciprocity.In recent years,it has been proved that indirect reciprocity plays a positive role in promoting cooperative behavior in groups,in which the reputation mechanism is an important way to realize indirect reciprocity.Individuals can gain a good reputation by providing help,so as to improve the probability of getting help from other individuals in the future.Some scholars have introduced adaptive reputation mechanism into the group.The level of reputation value will affect the transmission capacity of individual strategies,making the influence of reputation on cooperative behavior more direct.Based on the self-adaptive reputation mechanism,this paper conducts in-depth research,combines reputation mechanism with other mechanisms to classify groups,and explores the influence of individual classification on the evolution of group cooperative behavior.The main research contents and innovation points of this paper include the following two aspects:(1)Pure tag-mediated altruistic behavior is easy to disappear due to free-riding behavior,and cooperative behavior is not robust.Therefore,this paper introduces a tag and an adaptive reputation mechanism in the donation game on the spatial network.The tag is used to divide the individuals in the group into multiple sub-populations,and at the same time,each individual is given different behavioral attributes and random reputation value.In the study of this paper,each individual has five possible behavioral attributes,which are: nationalism,cosmopolitanism,intellectualism,egoism and altruism.In the process of game,players combine the three information of the other player's tag,reputation value and their own behavior attribute to make the final decision of game behavior,and then update the reputation value according to the individual's behavior.The simulation results show that after the introduction of reputation mechanism,the group cooperation rate can reach about85% in stable condition,which greatly promotes the improvement of cooperation level.(2)In the previous public goods games,it is assumed that individuals have the same resources,and only consider whether the investment is made or not,without involving different investment quantities,which is not consistent with the real situation.Therefore,this paper explores the evolutionary characteristics of group cooperation in public goods games with individual resource heterogeneity based on spatial network.In the group,the initial resource quantity distribution of individuals follows the power law distribution,and the group is divided into two types according to the number of resources it owns.Individuals with higher than the average resource of the group are individuals with high resource,and individuals with lower than the average resource are individuals with low resource.The two types of individuals have different investment quantities in the game.A large number of numerical simulations show that the participation of high-resource individuals makes cooperation more likely to occur.Reputation in the model can control the development of cooperation in the evolution process,but it does not affect the final result.