Modeling and control design of batch process plant by timed Petri net

Abstract

A design scheme for control of a batch process plant is proposed. The batch is assumed to be given in the form of events, conditions, their relations, and time, and is modeled by a timed Petri net. Analytical results are obtained based on the timed Petri net which describe the plant performance such as makespan, reachability state, and utilization of resources. The results are used in a plant controller which through dynamic optimization provides total control of the batch process plant. The resulting algorithm can be implemented either offline or online in real-time. The batch process design scheme discussed is illustrated by application to a small multipurpose batch process and implemented by means of a simulation package. It is noted that the results obtained through timed Petri net modeling, dynamic programming, and the theory of optimality can be used in designing the total real-time control of the batch process plant.<>