Controller performance of P,PI and neral network control in vinyl acetate monomer process.

This research is about investigating the controller performance between P, PI and Neural Network control in Vinyl Acetate Monomer (VAC) Process. The manufacturing process is about vapor-phase reaction converting ethylene (C2H4), oxygen (O2) and acetic acid (HAc) into vinyl acetate (VAc) with water (...

Full description

Saved in:
Bibliographic Details
Main Author: Faizal Nazareth, Ibrahim
Format: Undergraduates Project Papers
Language:English
Published: 2008
Subjects:
Online Access:http://umpir.ump.edu.my/id/eprint/672/1/19.Controller%20performance%20of%20P%2CPI%20and%20neral%20network%20control%20in%20vinyl%20acetate%20monomer%20process.pdf
http://umpir.ump.edu.my/id/eprint/672/
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:This research is about investigating the controller performance between P, PI and Neural Network control in Vinyl Acetate Monomer (VAC) Process. The manufacturing process is about vapor-phase reaction converting ethylene (C2H4), oxygen (O2) and acetic acid (HAc) into vinyl acetate (VAc) with water (H2O) and carbon dioxide (CO2) as byproducts. The data from the process are successfully generated and the simulation of the dynamic response is done with further analysis of P, PI control and Neural Network control. The study is focusing on the column section process as the clear view of the control performance is observed. The Proportional (P) and Proportional Integral (PI) control are type of controller that used in the process. The Neural Network control then is a control mechanism that has the similar system of human neurons for processing information data. It consists of network of neurons that have weight in each network and built generally in layers. As the analysis result of P and PI control showed some unsatisfying results, Neural Network Control is then developed to see the changes. In Neural Network control, the data has been trained and validate to get the better response before applied again to the process to see the improvement. At the end, Neural Network has visualized the better control performance as the unsatisfying responses of P and PI control have been improvised.