Abstract
A short-term schedule for oil refinery should arrange all the activities in every detail for the whole scheduling horizon, leading to a complex problem. There lacks efficient techniques and software tools for its solution applicable to industrial oil refinery. Considering that the feasibility of a schedule is essential, this paper studies the feasibility problem from a control perspective. A short-term schedule is composed of a series of operation decisions (ODs), each of which can be seen as a control. When an OD is executed, it transfers the system from one state to another. To guarantee the schedule feasibility, the system should be always be kept in safe states. The system is modeled by a Petri net model that is under control of the ODs. With this model, schedulability conditions for a system with one distiller are presented. These conditions reveal the relationship among the number of charging tanks, the oil transportation flow rate of the pipeline, and the production rate. The conditions are presented in a constructive way. Based on the conditions, when a realizable refining schedule is verified, a detailed short-term schedule is created for practical use.