Research On The Mechanisms Of Individual Adaptive Adjustment To Promote Cooperation In Spatial Prisoner's Dilemma

The evolution of cooperation is a significant scientific problem involving many disciplines.Cooperative behavior plays an important role in promoting the development of human society.However,cooperative behavior often weakens the advantage in the competition for survival.Understanding the widespread phenomenon of cooperation in nature and human society has long been an important and challenging issue.Evolutionary game theory on complex networks provides a powerful theoretical framework for explaining the emergence of cooperation,with game theory describing the interaction between individuals,and networks describing the spatial structure of systems.Many mechanisms have been studied on the evolution of cooperation such as spatial reciprocity,strategy-updating rules,behavioral diversity,etc.However,the role of adaptability in the dynamic game environment needs to be further explored and expanded.Therefore,this thesis focus on the influence of adaptability on cooperative evolution under the framework of the spatial repeated prisoner's dilemma game.The main contributions of this thesis are listed as follows.1.The influence of aspiration based interaction-range-adjustment on the evolution of cooperation is studied.Interaction-range is an important factor in the spatial prisoner's dilemma game.As traditional static interaction range does not adapt to the dynamic social environment,it is necessary to study the adaptive adjustment mechanism of the interaction range.In order to enhance the adaptability of individuals to the dynamic social environment,we propose an interaction range adjustment mechanism based on the aspiration of payoff.Simulation results show that the proposed mechanism can effectively promote the improvement of the level of cooperation.The probability of adjusting the interaction range has a great impact on cooperation evolution.The shrinking probability of interaction range is positively correlated with the probability that the system evolves into a fully cooperative state,while the increasing of expanding probability is not conducive to the emergence of cooperation.Moreover,the aspiration level is an important factor to affect the effectiveness of the interaction range adjustment.A higher level of aspiration can improve the conditions for individuals expanding their range of interaction,which reduces the risk of cooperators being exploited by defectors.The analysis of the underlying mechanism indicates that the‘isolation zone' formed in the early stage of evolution,which is composed of individuals with low interaction range,plays an important role in protecting the cooperator clusters.2.The influence of player-heterogeneity-based interaction-willingness-adjustment on cooperation evolution is studied.Player-heterogeneity level is proposed to measure the individual's ability to distinguish interaction partners.A high level of heterogeneity means that individuals can adjust their interaction-willingness with each other according to the results of the game.On the contrary,at a low level of heterogeneity,individuals will adjust their interaction-willingness according to the total payoffs of games,where interaction-willingness reflects the distance of individual social relationships.Simulation results show that the level of heterogeneity is not the higher the better,but is related to the way of interaction-willingness adjustment.Under the sensitive(large increase,large decrease)and radical(large increase,small decrease)adjustment pattern,a high level of heterogeneity is conducive to cooperation,while under the cautious(small increase,large decrease)adjustment pattern,a low level of player-heterogeneity is more conducive to cooperation.Treating each interaction partner differently affects the evolution of cooperation significantly.Especially when the interaction-willingness changes by a large step,the cooperation level and total payoff of the system can be improved significantly.In addition,a high level of heterogeneity can cut the interaction between cooperators and defectors which protects the cooperator clusters from being invaded.3.The influence of historical payoff-based interaction-willingness adjustment on cooperation evolution is studied.In the iterated prisoner's dilemma game,the memory length plays an important role in the adjustment of behavior.For individuals with memory,the long-term reward is more important than the immediate payoff.Hence,it is more realistic to take the historical payoff as a measuring factor of behavior adjustment.This thesis focuses on the role of memory length in the process of evolution and proposes an adaptive adjustment mechanism of interactionwillingness based on the historical payoff.The simulation results show that when the historical information is very short,the system can emerge cooperation under the relatively cautious(small increase,large decrease)adjustment of interaction-willingness,but cannot emerge cooperation under the stable(small increase,small decrease)and radical(large increase,small decrease)adjustment.The moderate length of historical payoff information can promote cooperation in the stable adjustment of interaction-willingness significantly.However,if the memory length was too long,cooperation can emerge only when interaction-willingness is slightly increased as needed.4.The influence of the hybrid strategy-updating mechanism on the evolution of cooperation is studied.Both social learning and self-reinforcement-learning have been extensively studied as promising mechanisms to promote cooperation in the context of spatial prisoner's dilemma.It is also widely recognized that both mechanisms are with deficiencies.In the purely social learning mechanism,individuals only imitate the strategies of other interaction partners,which makes the system unstable in a noisy game environment.In the purely self-learning mechanism,individuals are easy to become defectors because of self-interest.To compensate for the limitations of a single learning mechanism and promote the emergence of cooperation,this thesis proposes a hybrid learning mechanism.The results show that the proposed hybrid-learning mechanism facilities cooperation and has stronger evolutionary stability.In this mechanism,the weight of self-learning and the aspiration payoffs of individuals are two important factors that affect the level of cooperation.At a high aspiration level,self-learning is beneficial to the emergence of cooperation,while at a low aspiration level,social learning profits cooperation.