Data‐Augmented Deep Learning Algorithm for Accurate Control of Bioethanol Fermentation Using an Online Raman Analyzer

ABSTRACT Fed‐batch fermentation has become the preferred strategy in many industrial biomanufacturing processes. However, a key challenge remains in optimizing the feeding strategy to achieve stable maximum yields. In this study, we present an online Raman spectroscopy‐based monitoring and control s...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Biotechnology and bioengineering Jg. 122; H. 9; S. 2366 - 2376
Hauptverfasser: Ji, Kaidi, Yu, Xiaofei, Chen, Lifan, Wang, Yongbo, Guo, Zhiqiang, Chen, Biao, Li, Qingyang, Li, Zhen, Zhang, Hu, Wang, Guan, Zhuang, Yingping, Ruan, Yinlan
Format: Journal Article
Sprache:Englisch
Veröffentlicht: United States Wiley Subscription Services, Inc 01.09.2025
Schlagworte:
Algorithms
Artificial neural networks
Automation
Biofuels
Bioreactors - microbiology
Control systems
data augmentation
Deep Learning
Ethanol
Ethanol - analysis
Ethanol - metabolism
feedback control
Fermentation
Glucose
Labeling
Machine learning
Neural networks
Optimization
raman analyzer
Raman spectroscopy
Root-mean-square errors
Saccharomyces cerevisiae - metabolism
Semi-supervised learning
Spectroscopy
Spectrum analysis
Spectrum Analysis, Raman - methods
deep learning
fermentation
raman analyzer
data augmentation
feedback control
ISSN:0006-3592, 1097-0290, 1097-0290
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:ABSTRACT Fed‐batch fermentation has become the preferred strategy in many industrial biomanufacturing processes. However, a key challenge remains in optimizing the feeding strategy to achieve stable maximum yields. In this study, we present an online Raman spectroscopy‐based monitoring and control system, using bioethanol production by Saccharomyces cerevisiae as a case study. To address the issue of limited labeled data, a pseudo‐labeling approach based on semi‐supervised learning was employed, expanding the available training data set by 100‐fold compared to conventional labeling methods. In addition, we developed a spectral‐temporal concatenation convolutional neural network (STC‐CNN) that incorporates sequential spectral features. Comparative evaluations with multiple machine learning algorithms demonstrated the superior performance of STC‐CNN, achieving a root mean square error (RMSE) of 3.63 g/L for glucose prediction. The system enabled rapid and automated glucose feeding to maintain various target concentrations. Notably, a glucose setpoint of 30 g/L yielded the highest ethanol concentration of 140.68 g/L—an increase of 3.85% over traditional Fed‐batch fermentation—while reducing glycerol by 6.67%. These results highlight the significant potential of Raman spectroscopy combined with deep learning for automated bioprocess optimization and discovery of optimal operating strategies. A novel data‐augmented deep learning algorithm is introduced for real‐time control of bioethanol fermentation using an online Raman analyzer. By enhancing the predictive accuracy of fermentation parameters, this method significantly improves production efficiency, enabling optimized, autonomous fermentation processes and a substantial increase in bioethanol yield and sustainability.
Bibliographie:Kaidi Ji and Xiaofei Yu contributed equally to this study.
ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:0006-3592
1097-0290
1097-0290
DOI:10.1002/bit.29040