Modeling, optimization and control of discrete-event chemical processes using Petri net theory
Ekaterini C. Yamalidou · 1991
The purpose of this work is to use extended Petri net theory in order to model and analyze discrete-event chemical processes. The approach is to decompose the process into elementary units and to define the Petri net model for each such unit. The model of the larger process is constructed by connecting the elementary Petri net models according to interconnection rules, which depend on the scope of the study. Petri net models of several classes of discrete-event chemical processes are given as applications. These nets can be automatically generated since the modeling and interconnection rules are well defined. A way of translating the high-level Petri nets which model pipe/valve networks into constant matrices is described. Two simulation algorithms are developed, based on the Petri net model: one follows the evolution of the state of Petri nets with timed places and is used to compute the timing requirements, such as makespan, cycle time, equipment utilization times and product availability times, of the system under study; the other applies to Petri nets modeling pipe/valve networks and computes the system's state after a set of actions has been taken. Optimal control for the open-loop system is defined as an optimization problem whose set up is based on the Petri net model of the process. The constraints imposed by the function of the process are written as Boolean logic expressions and a way of translating them into linear inequalities is described. Finally, the problems of static control analysis and control synthesis for the closed-loop system are briefly discussed.