Artificial neural networks hold great promise for solving problems tha
t are extremely difficult to solve using conventional methods. We used
an artificial neural network to diagnose paper web breaks on a commer
cial newsprint paper machine. Process data for pulping and papermaking
operations were collected from the paper machine's distributed contro
l system. Additional data were obtained from on-line wet-end sensors (
zeta potential, first-pass retention, conductivity, and pH) that were
installed for this study. The essential variables contributing to pape
r web breaks were extracted using a three-stage multi-layer neural net
work and back-propagation method. Great savings of production costs we
re achieved: the number of web breaks was reduced, fiber loss in the e
ffluent was decreased, and workers spent less time cleaning, rethreadi
ng, and restarting the paper machine.