Co-evolution mechanism of peer effect and dynamic networks in multi-agent systems from a dual-game perspective

This research develops a multi-agent evolutionary game model that integrates dynamic networks, the peer effect strategy update rule, and a dual-game framework including the Prisoner's Dilemma Game (PDG) and the Snowdrift Game (SDG). Unlike previous research that analyzes these elements separately, we systematically examine their synergistic effect on the evolution of cooperation. Through simulations on the Newman–Watts (NW) small-world network and the Barabási–Albert (BA) scale-free network, we reveal a novel cooperation-emergence mechanism: the synergy between network structural adaptation and peer effect substantially accelerates the diffusion of cooperation and stabilizes network evolution. The results reveal that the adjustment of network topology exerts a more significant impact on improving system stability compared with the strategy update, while peer effect provides necessary social reinforcement for cooperative clustering. After stabilization, residual defectors are predominantly isolated nodes, and SDG participants exhibit a higher survival rate. These findings enrich the theoretical understanding of cooperative dynamics in complex systems and offer guidance for promoting cooperation in real-world multi-agent environments. • Build a dual-game model with dynamic networks and a peer-effect strategy update rule • Verify the model validity by simulations and between-group difference tests • Find the model reshapes network topologies and affects distinct networks differently • Peer effect and dynamic network structures collaborate to promote cooperation.