A generic simulation environment for agent-based FMS scheduling

Manufacturing scheduling is a classical optimization problem which is associated with high computational complexity causing low reactivity of conventional solutions. Scheduling for flexible manufacturing systems (FMSs) has to be performed dynamically with good reactivity and flexibility to adapt to the changing customer requirements. Agent-based scheduling is strongly advocated to tackle this challenge of combining the efficiency of the generated schedules with the flexibility of the scheduling process. The wide adoption of agent-based scheduling is hindered by the lack of simulation-based evaluation techniques. In this paper, a generic simulation environment for agent-based FMS scheduling is proposed. Details about the generation of dynamic FMS models and linking them to an agent-based scheduling framework, described in previous work, are given. By applying this generic simulation environment, the behavior of agent-based scheduling can be assessed for different shop floor control settings.