Neural Networks in Bioprocessing and Chemical Engineering

Baughman, D.R.; Liu, Y. A.

doi:10.1016/c2009-0-21189-5

Cited by 46 publications

(5 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Moreover, it is worth highlighting that the current hybrid modeling framework aims to provide a general structure for industrial bioprocess digitalization and optimization, and it can take advantages of different physical and datadriven modeling techniques to improve its capability. For instance, autoassociative neural network (AANN) can be used to replace the simple kinetic model for noise filtering (Baughman & Liu, 1995); current advances in machine learning based dynamic model structure discovery can be implemented at the top-level to identify the best physical model structure for process prediction and visualization;…”

Section: Discussionmentioning

confidence: 99%

“…The successful construction of data-driven models is strongly dependent on how the training data is formed to seek the best model structure. Hence, once high-quality data sets (replacing the original three sets) were generated at the framework middle-level, they were utilized to build a high-fidelity machine learning based ANN requires a large number of data sets to guarantee its interpolation power (Baughman & Liu, 1995), and its accuracy greatly depends on the selection of model hyperparameters (e.g., number of neurons, layers, and training epochs). Thus, several strategies developed in our recent research have been adopted here.…”

Section: Data-driven Model Constructionmentioning

confidence: 99%

“…Nonetheless, data-driven models are heavily reliant on the quality and quantity of data sets, currently limiting their application to industrial biosystems. Industrial data sets often contain large measurement errors and systematic noise, and it is not feasible to frequently measure all state variables (e.g., biomass concentration) at regular time intervals over the entire course of operation (Baughman & Liu, 1995). Second, most data-driven models calculate the values of state variables at fixed, pre-specified time intervals.…”

mentioning

confidence: 99%

“…Another particular challenge with data-driven models is that they are not based on physical mechanisms and are predominantly used to interpolate steady-state chemical processes (Baughman & Liu, 1995).…”

mentioning

confidence: 99%

“…Attention should be paid to the fact that ANN is not the only data-driven model that can be used in this framework. If the process is governed by multiple physical mechanisms, for example, biological kinetics and fluid dynamics, other models, for example, recurrent neural network can also be applied (Baughman & Liu, 1995).…”

mentioning

confidence: 99%

See 4 more Smart Citations

Hybrid physics‐based and data‐driven modeling for bioprocess online simulation and optimization

Zhang

Rio‐Chanona

Petsagkourakis

et al. 2019

Biotech & Bioengineering

View full text Add to dashboard Cite

Model‐based online optimization has not been widely applied to bioprocesses due to the challenges of modeling complex biological behaviors, low‐quality industrial measurements, and lack of visualization techniques for ongoing processes. This study proposes an innovative hybrid modeling framework which takes advantages of both physics‐based and data‐driven modeling for bioprocess online monitoring, prediction, and optimization. The framework initially generates high‐quality data by correcting raw process measurements via a physics‐based noise filter (a generally available simple kinetic model with high fitting but low predictive performance); then constructs a predictive data‐driven model to identify optimal control actions and predict discrete future bioprocess behaviors. Continuous future process trajectories are subsequently visualized by re‐fitting the simple kinetic model (soft sensor) using the data‐driven model predicted discrete future data points, enabling the accurate monitoring of ongoing processes at any operating time. This framework was tested to maximize fed‐batch microalgal lutein production by combining with different online optimization schemes and compared against the conventional open‐loop optimization technique. The optimal results using the proposed framework were found to be comparable to the theoretically best production, demonstrating its high predictive and flexible capabilities as well as its potential for industrial application.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Data-driven Model Constructionmentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

mentioning

confidence: 99%

See 3 more Smart Citations

Hybrid physics‐based and data‐driven modeling for bioprocess online simulation and optimization

Zhang

Rio‐Chanona

Petsagkourakis

et al. 2019

Biotech & Bioengineering

View full text Add to dashboard Cite

show abstract

Control of fermenters – a review

Rani¹,

Rao²

1999

Bioprocess Engineering

133

View full text Add to dashboard Cite

Fermenter control has been an active area of research and has attracted more attention in recent years. This is due to the new developments in other related areas which can be exploited to overcome the inherent dif®-culties in fermenter control. Beginning with conventional regulatory control of operating variables such as temperature, pH and dissolved oxygen concentration, research in fermenter control has undergone signi®cant changes including the recent neural network based approaches. The objective of the paper is to focus the attention of the researchers to the developments in the control of batch, fedbatch and continuous fermenters over the past few years.

show abstract