Resource state adaptive collaboration mechanism based on resource modeling and multi-agent system

The management of complex, dynamic, and cross-domain resources in cyber-physical-human systems (CPHS) faces significant challenges under spatiotemporal dynamics, particularly resource state conflicts caused by rapid environmental changes and interdependent resource interactions. To address these challenges, this study proposes an integrated framework combining resource modeling and resource state adaptive collaboration mechanism. First, a resource modeling framework for state coordination (RMFS) is developed to unify the representation of heterogeneous resources, their functionalities, and collaborative relationships through hybrid structural and semantic modeling. A resource state adaptive collaboration mechanism (RSACM) integrates multi-agent systems with knowledge graph to achieve real-time state synchronization. Agents utilize the collaborative relationships in the graph to make adaptive collaborative decisions on resource states, in order to perform state transitions and alleviate resource availability conflicts. Further, a meta-path-based resource inference (MPRI) method enables efficient resource retrieval and applies to simulation experiments by leveraging conceptual-instance meta-paths and large language model (LLM)-augmented substitution strategies to resolve resource unavailability. Experimental validation across emergency healthcare scenario demonstrates the framework’s effectiveness. An extension study was conducted on RMFS and RSACM through two cases from different fields. The proposed approach advances CPHS resource management by addressing heterogeneity, availability, and cooperativity in dynamic environments, offering theoretical and practical insights for complex system collaboration under spatiotemporal constraints.