A hybrid management method for heterogeneous non-cooperative behaviour in large-scale group decision-making
针对大规模群体决策中决策者因认知、价值观等差异产生的异质性非合作行为，提出一种结合自适应聚类、偏好结构共识测量和最小调整反馈机制的混合管理方法，并通过数值实验验证其有效性。
Driven by technological and societal demands, decision science has rapidly advanced, with large-scale group decision-making (LSGDM) widely applied to complex problems across diverse fields. In LSGDM, non-cooperative behaviours of decision-makers (DMs) often hinder consensus-reaching, making their effective identification and management critical for improving decision-making efficiency. Given differences among DMs in cognitive patterns, value orientations, and risk preferences, the motivations behind their non-cooperative behaviours exhibit significant heterogeneity. Therefore, a novel LSGDM framework considering heterogeneous non-cooperative behaviours is proposed. Specifically, an adaptive hierarchical clustering algorithm that integrates cohesion and opinion diversity is proposed to form subgroups that foster collaboration while preserving opinion diversity. Furthermore, a consensus measure method based on preference structures is proposed to accurately reflect the group consensus level. To flexibly provide adjustments for DMs, a feedback mechanism based on the minimum adjustment consensus model (MACM) is formulated, incorporating three types of feedback strategies. From the perspective of heterogeneous motivations underlying non-cooperative behaviours, a novel method is proposed to identify them. Following a “guidance-oriented with intervention as a supplement” principle, a hybrid management method is introduced to address these behaviours. Finally, a numerical example demonstrates the feasibility of the method, while comparative and simulation analyses highlight its advantages and effectiveness.