A Hybrid Modeling for Continuous Casting Scheduling Problem

This paper deals with a multi-agent-based interval type-2 fuzzy (IT2F) expert system for scheduling steel continuous casting. Continuous caster scheduling is a complex and extensive process that needs expert staff. In this study, a distributed multi-agent-based structure is proposed as a solution. The agents used herein can cooperate with each other via various communication protocols. To facilitate such communication, an appropriate negotiation protocol (i.e., contract net protocol) is proposed. The due dates specified by expert staff are represented by IT2F membership functions (MFs). As a part of the objective functions, a simple procedure is proposed to calculate the total earliness and tardiness penalty when the due date’s MFs are IT2F. The proposed hybrid multi-agentbased system combines the multi-agent systems with type-2 fuzzy concepts which conforms to the real-world continuous casting problem.