Neural Network Based Multi Stage Modelling of Chylla Haase Polymerization Reactor
Author:
Damodaran Vasanthi,N Pappa
Abstract
Abstract
An accurate semi batch process model should be a nonlinear dynamic model. Neural networks are suitable for modelling nonlinear dynamics and can be used for developing empirical models of semi batch processes. Multi stage neural network based modelling of the polymerization reactor described by Chylla and Haase, is illustrated in this paper. The process is divided into three regions namely heat up period, feed period and hold period and neural model is developed for each stage. This method of multi stage modelling captures the dynamics of the process accurately for the semi batch process. At different stages respective neural model is active based on the period of operation.
Publisher
Walter de Gruyter GmbH
Subject
Modeling and Simulation,General Chemical Engineering