Decentralized production planning using collaborative agents for a bi-objective reeentrant hybrid flow shop

Abstract

In this paper, we consider production planning in a reentrant hybrid flow shop with serial stages. Each stage comprises of identical parallel machines and products can be processed on any one of these machines at any stage. Also, a product may revisit some stages multiple times. We consider a Thin Film Transistor and Liquid Crystal Display (TFT-LCD) industry production planning problem and propose a two-phase decentralized coordination mechanism that allows entities to partake in planning decisions while preserving their autonomy to achieve a globally feasible plan. Also, we provide computational results of the application of the proposed mechanism to randomly generated problems in comparison to other known methods in literature in terms of WIP, delayed demand and throughput.