Manipulate holistic cooperation of population game:optimal induction strategy for a few irrational agents in social networks  

作　　者：Zhen WANG Peixuan SONG Da-Tian PENG Dengxiu YU 

机构地区：[1]School of Cybersecurity,Northwestern Polytechnical University,Xi’an 710072,China

出　　处：《Science China(Information Sciences)》2025年第4期332-344,共13页中国科学(信息科学)(英文版)

基　　金：supported in part by National Science Fund for Distinguished Young Scholars(Grant No.62025602);National Natural Science Foundation of China(Grant Nos.U22B2036,62373302);Fundamental Research Funds for the Central Universities(Grant Nos.G2024WD0151,D5000240309);Tencent Foundation;XPLORER PRIZE。

摘　　要：In social networks,cooperation dilemmas among agents persist as a significant challenge.Existing research often emphasizes mild interactions among rational agents for autonomous cooperation,but it tends to overlook the impact of irrational agents'decisions,particularly under intentional inductions.This paper proposes an optimal induction strategy to enhance overall cooperation within the network,even when only one irrational agent is involved.We employ a continuous action iterative prisoner dilemma framework to model multi-agent games and utilize a tracker to minimize cooperation costs.A heuristic regulation algorithm is employed to generate optimal induction strategies,which may involve selecting a rational agent as a proxy for the irrational one.Through theoretical analysis using the Lyapunov direct method,we demonstrate the convergence of game dynamics despite the presence of irrational agents.Numerical results validate the effectiveness of our proposed strategy in promoting holistic cooperation.These findings highlight the vulnerability of social networks to external incentives and pave the way for further research into optimizing multi-agent cooperation dynamics.

关 键 词：population game prisoner’s dilemma convergence cooperation tightness social network 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]