Continuous Time Modeling for Short Term Scheduling of Multipurpose Pipeless Plants

Abstract

An efficient short-term scheduling, mixed integer programming model for a multipurpose pipeless plant over a continuous-time domain is addressed. In contrast to the conventional scheduling models for the pipeless plants with discrete time representation requiring a large number of 0-1 variables, the continuous-time model where each product has deterministic processing stage blocks with duration for unit allocation and each processing unit has a corresponding time slot is built. Jobshop features such as re-entrant production flows and diverse processing directions are effectively modeled by identifying the stage number of products in an invariant way regardless of processing sequence or recipe. The performances of the model and the solution method are illustrated through two examples.