Multi-agent game theory and collaborative mechanisms in smart supply chain scenarios

This paper proposes a multi-agent collaborative enhanced supply chain management (MACE-SCM) framework that integrates cooperative game theory, intelligent process computation, and hybrid communication protocols to achieve adaptive coordination in complex supply chain networks. Each operational entity—supplier, wholesaler, retailer, logistics, and third-party logistics—is modeled as a hierarchical agent consisting of communication, collaboration, and control layers. The AG six-factor vectorization model is introduced to represent workflow actions as computational gene vectors, enabling precise responsibility mapping and AI feasibility evaluation. A dual-layer communication mechanism combining contract net protocol (CNP) and HTTP/REST ensures real-time negotiation and feedback among agents. Simulation experiments demonstrate that MACE-SCM observable improves order completion rate, profitability, and system stability compared with conventional models, even under uncertain demand and capacity conditions. The results validate the proposed framework’s effectiveness in enhancing intelligence maturity, operational resilience, and global optimization of supply chain systems.