Continuous time formulation of short-term scheduling for pipeless batch plants

Abstract

An efficient scheduling of pipeless plants is proposed as mixed integer linear programming models. As previous models based on uniform time discretization methods or two-coordinate time representation methods cause a large number of binary variables, a continuous time representation model based on conventional multipurpose/multiproduct formulations is proposed. It is composed of three formulations, which are the production-restricted scheduling, the vessel-restricted scheduling, and the waiting-station-restricted scheduling. When the number of products increases, the size of a proposed formulation exponentially increases. Therefore, periodic scheduling is considered and an iterative solution procedure is proposed. The performance of the proposed models and the solution procedures are illustrated with several examples.