This solution is designed to reduce starch consumption in packaging paper production while maintaining the required paper strength and quality. It combines a virtual sensor for online SCT value prediction with an advisory model for starch optimization on the size press. The system was developed using historical DCS data and laboratory SCT results, with key variables such as grammage, moisture, and specific starch consumption used to build the prediction model.
The selected model was implemented as a stand-alone Windows Service application that acquires data from the DCS through the ERGA OPC Client, predicts SCT values, and sends the results back to the DCS with minimal changes to the existing automation system. This allows predicted SCT values to be displayed in SCADA and included in production reports.
By combining machine learning, virtual sensing, and process optimization, the solution helps paper producers lower raw material costs, improve efficiency, and maintain stable product quality in real production conditions.
Hotline:
+49 (0)511 89-0
Deutsche Messe AG
Cookie Settings