Print-through prediction using ANNs

Kesheng Wang and Berit Lorentzen

A newsprint mill is using modeling techniques to control print-through on the paper machine. The
model analyses the key variables in process and furnish which affect print-through, and thus enables quality
teams to predict print-through and make the necessary adjustments to the furnish or the process. In this paper, we
use Artificial Neural Networks (ANNs) modeling approach to control print-through level instead of regression
models. The intelligent model is flexible, noise-resistant, fast and reliable. The model is developed by a
commercial Artificial Neural Network development tool - Neuframe. Relative merits of ANNs for the automatic
visual inspection task have been discussed. The system developed can be used for quality control and production
management of paper products.

NOTE: TO READ THIS AND OTHER IPROMS 2006 PAPERS, PLEASE REGISTER FOR THE CONFERENCE.

REGISTRATION IS FREE.

CLICK here TO REGISTER.

Attachment	Size
Print-through paper PPT_001-1.wmv	17.74 MB

Submitted by soroka on Sun, 02/07/2006 - 12:21pm.
login or register to post comments |

A question about literature review

Submitted by LiuH on Mon, 03/07/2006 - 7:35pm.

In the paper, you use Artificial Neural Networks (ANNs) modeling approach to predicate print through. Have you thought of solving this particular problem by using other machine learning methods and comparing them? Thank you.

Dear authors, Thanks for

Submitted by LiuH on Wed, 12/07/2006 - 9:09am.

Dear authors,

Thanks for your contribution to IPROMS 2006. If you happen to see this message, please respond the query ASAP. Thank you very much for your cooperation.

Submitted by Wangk on Thu, 13/07/2006 - 12:41pm.

We have tried Fuzzy logic system and Neural-fuzzy system for pridiction of print-through of the paper as well. We feel that the ANN model is the best selection because it is much easy and reliable!

Print-through prediction using ANNs

Comment viewing options

User login