Skip to Main Content
A single supply chain configuration will neither be optimal nor efficient under dynamic and uncertain conditions, where objectives may conflict. Thus, the issue of dynamic configuration of supply chains needs serious research attention. In this paper, we combine a high level petri net with probabilistic reasoning as probabilistic petri nets to model a Multi-agent system (MAS) and detecting together goal and plan conflicts dynamically and concurrently for supply chain networks. We model supply chain dynamics based on conflicts. An integrated framework to tackle conflicts includes two stages: Conflict Recognition (CR) and Conflict Investigation (CI). CR module acts as a monitor to dynamically detect different conflicts; once conflicts identified, conflict investigation module is triggered to evaluate and rank the conflicts, which indicates to what extent the conflict will impact the agent's behavior. The model explicitly captures the interactions among supply chains and within supply chains.