Image‐Based Flow Regime Recognition in Aerated Stirred Tanks Using Deep Transfer Learning

Abstract

Monitoring of flow regimes in aerated stirred tanks is important to ensure energy efficiency and product quality. The use of deep learning models for the recognition of flow regimes shows promising results. However, such models require a large amount of data for training. The aim of this paper is to apply the deep transfer learning approach to address this challenge. We compare various pre‐trained models with the differential learning rate and 2‐step transfer learning approaches to analyse the resultant model performance. We also investigate the effect of the dataset size on the classification accuracy.

AI in Process Industries – Current Status and Future Prospects

“…Flow regimes in aerated stirred tanks are in general good indicators for final quality and efficiency. Their imagebased classification is analyzed in [57] using deep learning based on different pre-trained models to reduce data need.…”

70 Jahre Verfahrenstechnik an der TU Dresden

“…Prof. Urbas koordiniert das Verbundprojekt KEEN zu diesem Thema. Im Rahmen dieses Projekts wurden Methoden zur Erkennung von Stro ¨mungsregimes in begasten Beha ¨ltern mittels Bilddaten erforscht [94], Werkzeuge zur Integration derartiger komplexer Datenquellen in modulare Anlagen entwickelt [95] und eine Methode zur automatischen Evaluation von biochemischen KPIs auf Basis von DEXPI-Informationen erarbeitet [96]. Die Erfahrung zeigt, dass die großen Datenmengen, die in der Prozessindustrie vorhanden sind, ha ¨ufig leider nur wenig Information enthalten.…”