Scheduling of process plants with equipment changeover

Abstract

Changeovers are necessary in a manufacturing environment, however effective equipment time is lost in cleaning or setting up the equipment for a different product or recipe. Efficient modelling of these changeovers is required to obtain optimal schedules to minimize the losses. For this a Mixed Integer Linear Programming (MILP) approach is presented to optimally schedule such changeovers. In this time dependent and frequency dependent changeovers are initially presented. Then a formulation to schedule general degradation model dependent changeover is developed. A binary variable to denote global precedence of processing of an order with respect to time of changeovers (cleaning) is introduced in the formulation. The formulation is first applied to a general flowshop to study its features. It is then applied to an ice-cream production facility with multiple stages and parallel equipment to demonstrate its wider applicability. The approach is validated with results available in the literature and in one of the cases found a marginally better schedule.