Quality Monitoring through a Dynamic Neural software sensor

Description

The paper discusses the application of a software sensor to the on-line monitoring of a terpolymerization reactor. The monitored quality indicator is the Mooney viscosity that is one of the main quality indicators used to grade the final product. We will illustrate the results obtained through the off-line implementation of a dynamic neural sensor whose performance is also compared with that of a previous sensor based on feed-forward neural networks.

Additional details