Generative Artificial Intelligence-Guided Quantum Group Consensus Based on Incentive Behavioral Optimization under Multi-Scale Environments for Consumer Electronics in Smart Cities

Rui Li, Chao Zhang, Arun Kumar Sangaiah, Jaya Subalakshmi Ramamoorthi, Turki Alghamdi, Majed Aborokbah

Abstract: In the development of smart cities, consumer electronics, especially cameras, and generative artificial intelligence (GAI), especially large language models (LLMs), are being integrated into agricultural disease prevention. To prevent wheat stripe rust (WSR), an agricultural disease, decision-makers (DMs) need to evaluate disease conditions across different regions and reach a consensus to scientifically select optimal wheat planting regions. However, current group consensus methods struggle with representing multi-source information and lack personalized incentives to engage diverse DMs. Therefore, an LLM-guided group consensus method for the location selection of wheat planting regions (LS-WPR) under structured multi-scale information systems (MSISs) is explored. First, ChatGPT-4o converts wheat images into numerical forms, which is used to build an MSIS by hierarchical clustering and select optimal scale combinations by three-way decisions (TWD). Second, trust weights are calculated via McKinsey trust equation, and quantum-like Bayesian networks (QBNs) model interference among DMs to refine their weights. Third, DMs are classified based on their behavioral characteristics using three-way clustering (TWC), and a personalized behavior optimization mechanism with trust-driven incentive strategies is applied to achieve LS-WPR consensus. Fourth, quantum frameworks are used to maximize differentiation among regions. Finally, the method is validated by experiments to demonstrate its advantages in LS-WPR.

External IDs:doi:10.1109/tce.2025.3624313