An Agent-Based Architecture for Solving Dynamic Ressource Allocation Problems in Manufacturing

Abstract

In this paper, we present results of agent-based modeling for the solution of dynamic resource allocation problems in manufacturing. After a concise problem description, we start by identifying the agents necessary for the required shop floor control functionality. Then we describe the architecture and implementation of a multi-agent-system prototype based on the object-oriented framework JAFMAS, a blackboard-type data layer and the discrete event simulator ParSimONY. We present results of computational experiments with our prototype multi-agent-system.