A multi-objective optimization consensus model for large-scale group decision-making considering dynamic social networks

As global climate change intensifies and fossil fuel resources continue to deplete, transitioning to alternative energy sources has become a vital strategy for sustainable development. Photovoltaic (PV) power generation, recognized for its clean, renewable, and low-carbon characteristics, is advancing rapidly. Against this background, site selection for PV power plant stations has become a crucial decision-making factor in ensuring project success. However, in large-scale group decision-making (LGDM), the complex backgrounds and substantial number of decision-makers (DMs) pose a significant challenge in reaching consensus efficiently. Therefore, this paper proposes a multi-objective optimization consensus model (MOOCM) utilizing dynamic trust networks to solve LGDM problems. First, a hybrid trust network (HTN) is built by integrating preference similarity and trust relationships, and DMs are clustered using the Louvain algorithm based on this hybrid network. Second, a MOOCM is designed with the objectives of minimizing costs, maximizing fairness, and achieving a high consensus level. Then, after consensus is reached, the HTN is updated, and secondary clustering is performed to obtain dynamic weights for DMs. Finally, a PV power plant site selection problem with 20 DMs, four alternatives, and four attributes is used as a case study for validation. In the first clustering, the DMs are divided into four subgroups. After consensus is reached, the HTN is updated and a second clustering is performed, which finally produces three subgroups. At the same time, the proposed method can achieve a group consensus level (GCL) of 0.9597, with Cost = 1.7422 and Fairness = 0.9043. These results verify the effectiveness and practical utility of the proposed method in LGDM.