2022
DOI: 10.1021/acs.iecr.1c04534
|View full text |Cite
|
Sign up to set email alerts
|

Design and Optimization of a Penicillin Fed-Batch Reactor Based on a Deep Learning Fault Detection and Diagnostic Model

Abstract: The application of a supervised deep convolutional autoencoder was tested against partial least-squares-discriminant analysis (PLS-DA) for fault detection and diagnosis in a penicillin fed-batch process. In silico data was generated with a comprehensive simulator (IndPenSim) of an industrial-scale penicillin fed-batch simulator of operation under normal batch conditions and 8 fault batch conditions. A composite face-centered design response surface was applied to optimize key bioreactor design parameters based… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
3
0

Year Published

2022
2022
2025
2025

Publication Types

Select...
9

Relationship

1
8

Authors

Journals

citations
Cited by 10 publications
(5 citation statements)
references
References 24 publications
0
3
0
Order By: Relevance
“…In this paper, a state-of-the-art penicillin fermentation simulation platform, Pensim V 2.0 [40,41], is employed to emphasize the advantages of CDVAE.…”
Section: Process Description and Modelingmentioning
confidence: 99%
“…In this paper, a state-of-the-art penicillin fermentation simulation platform, Pensim V 2.0 [40,41], is employed to emphasize the advantages of CDVAE.…”
Section: Process Description and Modelingmentioning
confidence: 99%
“…For example, Hematillake, D and others put forward an automatic sketch generation method based on CNN. By learning a large number of design samples, the model can generate a preliminary sketch that conforms to the design semantics [7] . However, there is still room for improvement in details and diversity of these methods.…”
Section: ⅱ Related Workmentioning
confidence: 99%
“…The hyper-parameter search is implemented using a Keras-tuner. Firstly, a grid of hyper-parameters is defined, for example the number of encoder layers = [1,2,3], number of LSTM units for each of these layers ranging from 2 to 200 with an interval of 2 = [10:2:200], learning rate = [0.1, 0.2, 0.3, 0.01], value of weights in the objective function, etc. Keras-tuner trains the model using different combinations of these hyper-parameters values, and the averaged validation accuracy is evaluated at every epoch.…”
Section: Counts Of Predicted Label I Counts Of Predicted Label Other ...mentioning
confidence: 99%
“…The operation of industrial plants employs sensors and control loops to mitigate the economic losses resulting from these faults. However, in the presence of process faults and manipulated variable constraints, these control schemes are not sufficiently resilient to avoid abnormal operation [1,2]. Thus, process faults must be diagnosed and addressed by implementing a suitable corrective measure.…”
Section: Introductionmentioning
confidence: 99%