Consensus modeling for maximum expert with quadratic cost under various uncertain contexts: A data-driven robust approach

Consensus optimization models are valuable tools for addressing negotiated group decision-making challenges, particularly those involving critical decision-related data such as costs and preferences. However, the idealized approach to information fusion in consensus decision-making presents challenges in adapting to practical conditions, leading to less credible consensus solutions. To simulate a more realistic decision-making scenario, this study integrates unit adjustment costs in a quadratic form into a maximum expert consensus model. This quadratic cost formulation captures the complex resistance to cost changes encountered by experts when adjusting solutions, promoting a deliberate approach to solution updates and facilitating improved decision-making. Moreover, economic insights elucidate the effect of quadratic costs on decision-making behavior. Additionally, the feasibility of reaching a consensus may be impeded by high uncertainty in real-world decision-making scenarios. This study separately tackles decision environments characterized by unit adjustment costs and individual preference uncertainty. It employs a robust optimization approach to incorporate uncertain costs and preferences into the optimization model. Data-driven robust maximum expert consensus models are then developed to objectively manage available historical data. An enhanced genetic algorithm is introduced as a solution method to address the proposed models. The proposed models are ultimately applied to evaluate policy options for the development of new energy vehicles in Changsha. Comparative and sensitivity analyses are conducted, showing the superior performance of the proposed models.