Optimization of recipe based batch control systems using neural networks

In the modern pharmaceutical industry many flexible batch plants operate under an integrated business and production system, using ISA S95 and ISA S88 standards for models and terminology, and implementing flexible recipe-based production. In the environment of constantly changing market conditions,...

Full description

Saved in:
Bibliographic Details
Published in:Chemical and biochemical engineering quarterly 2012-09, Vol.26 (3), p.175
Main Authors: Sostarec, A, Gosak, D, Hlupic, N
Format: Article
Language:English
Subjects:
Control systems
Neural networks
Production processes
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In the modern pharmaceutical industry many flexible batch plants operate under an integrated business and production system, using ISA S95 and ISA S88 standards for models and terminology, and implementing flexible recipe-based production. In the environment of constantly changing market conditions, adjustment to surroundings is a business necessity. To support necessary production improvement, regulatory authorities have introduced the risk based approach for the control of process development, production based on the quality by design (QbD) principle, and process analytical technology (PAT). In this work, the method for practical implementation of an adaptable control recipe, that allows process improvement inside the previously established design space, is proposed, based on the neural network process model. Based on the neural network model, the three methods for recipe-controlled process improvement and optimization were introduced - neural-based software sensor, generic neural model control, and process optimization using iterative dynamic programming. Suitability of the proposed method was tested in a mini reaction plant Chemreactor Buchi, running the wastewater treatment batch, controlled by the production recipe based on S88 standard. Key words: ISA S88, neural network model, recipe-controlled process, design space
ISSN:0352-9568