Evolutionary Games And Opinion Dynamics In Complex Networks

With the rapid development of network science,the complex network has become an important method to describe the interaction between individuals within a complex system.We can analyze many complex systems in real life by modeling them on complex networks.On the other hand,by investigating various evolutionary dynamics on complex networks,people can deeply understand many complex phenomena existing in nature and real social systems,such as the spontaneous emergence of cooperative behavior among selfish individuals,the formation and spread of the opinion on social networks,the synchronous phenomenon that exists widely in the nature.Therefore,as important research areas in the field of complex network,evolutionary dynamics on complex networks have received extensive attention and research,such as evolutionary games,opinion dynamics and the dynamics of coupled oscillators,and so on.In this dissertation,based on evolutionary games and opinion dynamics on complex networks,we have investigated the evolution of cooperation and the consensus of opinion on complex networks.The main contents of this dissertation are summarized as follows:For the evolutionary game dynamics on complex networks,(1)we have studied the evolutionary prisoner's dilemma game with the introduction of strategy persistence.We find that persistence could promote cooperation in a population no matter what the network structure is.Furthermore,the results on heterogeneous networks show that individuals with larger persistence level are more inclined to cooperate than those with smaller persistence level.Moreover,we find that higher cooperation levels can be reached if the leaders(the individuals with high degrees)persist their strategies regardless of the structures of the underlying networks.(2)We introduce a preferential selection mechanism based on degree difference into spatial prisoner's dilemma games on ER random networks and BA scale-free networks and investigate the effects of the preferential selection on cooperation.The results show that,when the individuals prefer to choose the neighbors who have small degree difference with themselves to imitate,cooperation is hurt by the preferential selection.In contrast,when the individuals prefer to choose those large degree difference neighbors to learn from,there exists an optimal preference strength resulting in the maximal cooperation level no matter what the network structure is.(3)We consider the spatial public goods game on two-layer networks,where the interaction and learning environments are represented by two asymmetric layers,respectively.We focus on the effects of edge overlap ? between the interaction and learning networks on the evolution of cooperation.The simulation results show that,the effects of ?on the evolution of cooperation depend on the synergy factor r.For relatively small r,higher overlap between the interaction and learning environments is more favorable for cooperation.Contrarily,for relatively large r,the lower overlap between the interaction and learning environments results in higher level of cooperation.For the opinion dynamics on complex networks,we study the effects of heterogeneous convergence rate on consensus in Deffuant model.The convergence rate depends on the opinion difference between interacting agents and a tunable parameter ?.For nonzero ?,both a large opinion difference and a large ? can slow down the convergence of opinions between the interacting agents.The simulation results show that the increase of ? is favorable to decreasing the confidence threshold for the population to always reach complete consensus,and there exists optimal ? resulting in the minimal bounded confidence threshold.It means that appropriately decreasing the convergence rate in each interaction is favorable for consensus.