Opinion Dynamics Based on Game Theory in Multi-agent Network

With the development of multi-agent systems including swarm robot systems, individual agents' decisions need to be propagated in the networks at a higher speed and coverage. Therefore, analyzing the decision-making changes of the individuals and groups toward events after interactions under environmental influence is important, and it could be helped by the research of opinion dynamics, i.e. the opinion formation process. Also, the interactions between agents which hold different decisions in the opinion formation model can be constructed based on game theory. In this work, the characteristics and disciplines of the opinion formation process are analyzed by applying nonrational game models. Furthermore, a memory mechanism is proposed to improve the game model. Finally, based on the obtained model, a simulation of the opinion formation process is carried out, and the results are systematically analyzed.