Research And Control Of Multiple Individuals Collaboration Under Dynamic Evolution Game Mechanism

Mechanism of evolutionary game theory is a combination of evolutionary theory and analysis of the dynamic evolution.It can provide a good explanation of methodology for group collaboration and effectively solve local deviation phenomena in multi-individual control system.On the basis of the above,this paper focuses on multi-inividual cooperation and control within the framework of evolutionary game.The main work are as follow:Firstly,the development and related concepts of evolutionary game is briefly summarized.The principles and inherent laws of classical game models are analyzed with corresponding experiments.Besides,the common tragedy problem in public goods game is discussed.It demonstrated that the improved punishment method is efficient in solving such a problem and enhancing the cooperation level.Secondly,based on the background of mutated multi-individual coordination mode,this paper takes multi-individual flocking movement system as an example.The concept of mutated strategy is proposed,and two dynamic performance indicators including convergence time and stabilization time are defined to contrast the dynamic performances of the corresponding flocking system after mutation appeared,which is a innovation point.Then the system is optimized by voluntarily memorizing the optimal strategy according to update equation under self-learning control mechanism.The experimental results demonstrated that the system could gradually achieve performance optimization at some degree.But it failed to achieve the goal of evolution consistency.Finally,to rapidly resume strategy consistency in mutated flocking system,the evolutionary game mechanism is utilized for contrast experiments.To establish direct links between payoff matrix and dynamic performance indicators,the payoff matrix parameters are quantified with the two dynamic performance indicators in this paper,which is another innovation point in this paper.By adjusting payoff matrix parameters of micro individuals,various evolution results could be accomplished,including original strategy consistency,mutated strategy consistency and mixed strategy coexistence.Meanwhile the dynamic performances like convergence time and stable time could be greatly enhanced.