Identification of semi-parametric hybrid process models

Yang, Aidong; Martin, E.B.; Morris, A.J.

doi:10.1016/j.compchemeng.2010.05.002

Cited by 18 publications

(15 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Before starting to identify model parameters, it is important to identify and remove outliers. Outliers can increase the level of variance of the model parameters (Yang et al 2011 ), can reduce the model performance by biasing parameter estimates, and can lead to false conclusion. Outliers are often due to fault, biological deviations, or human/instrumental errors.…”

Section: Model Calibration and Testingmentioning

confidence: 99%

Macroscopic modeling of mammalian cell growth and metabolism

Yahia

Malphettes

Heinzle

2015

Appl Microbiol Biotechnol

View full text Add to dashboard Cite

We review major modeling strategies and methods to understand and simulate the macroscopic behavior of mammalian cells. These strategies comprise two important steps: the first step is to identify stoichiometric relationships for the cultured cells connecting the extracellular inputs and outputs. In a second step, macroscopic kinetic models are introduced. These relationships together with bioreactor and metabolite balances provide a complete description of a system in the form of a set of differential equations. These can be used for the simulation of cell culture performance and further for optimization of production.

show abstract

Section: Model Calibration and Testingmentioning

confidence: 99%

Macroscopic modeling of mammalian cell growth and metabolism

Yahia

Malphettes

Heinzle

2015

Appl Microbiol Biotechnol

View full text Add to dashboard Cite

show abstract

“…Integration is performed for the time span of a day and eventually, the feed addition is simulated as follows:

C_{i, normali normaln normali normalt} (T) = C_{normali} (T) + \frac{normalF normale normale d_{i} (T)}{V_{R}},

where C i ( T ) represents the concentration value before the feed addition, which also corresponds to the measured value, while C i , init ( T ) represents the initial condition for integrating Equation from T to T + 1, Feed i ( T ) the mass of species i , that is fed, and V R the reactor volume. A two‐norm regularized objective function (Yang et al, ) is used to avoid overfitting of the weights in the neural network and five‐fold cross‐validation (Hastie, Tibsharani, & Friedman, ) is used to determine the optimal number of nodes and the regularization parameter. ode15s ( ) is used for integration and fminunc ( ) is used for optimization purposes.…”

Section: Methodsmentioning

confidence: 99%

“…Differences among these applications relate not only to the model architecture but also to the algorithm used in the data‐driven parts. Although in most cases artificial neural networks (ANNs) were used some applications of support vector regression (Hu et al, ; Yang, Martin, & Morris, ), nonlinear partial least square (PLS; Von Stosch, Oliveira, Peres, & Feyo de Azevedo, ), and other Black‐box models (Tian et al, ) can also be found.…”

Section: Introductionmentioning

confidence: 99%

A new generation of predictive models: The added value of hybrid models for manufacturing processes of therapeutic proteins

Narayanan

Sokolov

Morbidelli

et al. 2019

Biotech & Bioengineering

View full text Add to dashboard Cite

Due to the lack of complete understanding of metabolic networks and reaction pathways, establishing a universal mechanistic model for mammalian cell culture processes remains a challenge. Contrarily, data-driven approaches for modeling these processes lack extrapolation capabilities. Hybrid modeling is a technique that exploits the synergy between the two modeling methods. Although mammalian cell cultures are among the most relevant processes in biotechnology and indeed looks ideal for hybrid modeling, their application has only been proposed but never developed in the literature. This study provides a quantitative assessment of the improvement brought by hybrid models with respect to the state-of-the-art statistical predictive models in the context of therapeutic protein production. This is illustrated using a dataset obtained from a 3.5 L fed-batch experiment. With the goal to robustly define the process design space, hybrid models reveal a superior capability to predict the time evolution of different process variables using only the initial and process conditions in comparison to the statistical models. Hybrid models not only feature more accurate prediction results but also demonstrate better robustness and extrapolation capabilities. For the future application, this study highlights the added value of hybrid modeling for model-based process optimization and design of experiments. K E Y W O R D Sbiopharmaceuticals, hybrid models, mammalian culture, predictive modeling, process optimization

show abstract

“…Most of the hybrid modeling work pose basic mass or energy balances and some preliminary dependencies and approximate unknown relationships with data-driven models. The most popular data-driven model used in such frameworks is a shallow (typically single layer) artificial neural network [6,8,9,[13][14][15][16] with some literature also reporting use of subspace identification algorithms [17], non-linear Partial Least Squares [18], Adaptive regression splines [19], Support Vector Machine (for regression) [20,21], Gaussian Processes [22] and Recurrent Neural Network [7].…”

Section: Introductionmentioning

confidence: 99%

Functional-Hybrid modeling through automated adaptive symbolic regression for interpretable mathematical expressions

Narayanan

Bournazou²,

Gosálbez

et al. 2022

Chemical Engineering Journal

View full text Add to dashboard Cite

Mathematical models used for the representation of (bio)-chemical processes can be grouped into two broad paradigms: white-box or mechanistic models, completely based on knowledge or black-box data-driven models based on patterns observed in data. However, in the past twodecade, hybrid modeling that explores the synergy between the two paradigms has emerged as a pragmatic compromise. The data-driven part of these have been largely based on conventional machine learning algorithm (e.g., artificial neural network, support vector regression), which prevents interpretability of the finally learnt model by the domain-experts.In this work we present a novel hybrid modeling framework, the Functional-Hybrid model, that uses the ranked domain-specific functional beliefs together with symbolic regression to develop dynamic models. We demonstrate the successful implementation of these hybrid models for four benchmark systems and a microbial fermentation reactor, all of which are systems of (bio)chemical relevance. We also demonstrate that compared to a similar implementation with the conventional ANN, the performance of Functional-Hybrid model is at least two times better in interpolation and extrapolation. Additionally, the proposed framework can learn the dynamics in 50% lower number of experiments. This improved performance can be attributed to the structure imposed by the functional transformations introduced in the Functional-Hybrid model.

show abstract

Identification of semi-parametric hybrid process models

Cited by 18 publications

References 28 publications

Macroscopic modeling of mammalian cell growth and metabolism

Macroscopic modeling of mammalian cell growth and metabolism

A new generation of predictive models: The added value of hybrid models for manufacturing processes of therapeutic proteins

Functional-Hybrid modeling through automated adaptive symbolic regression for interpretable mathematical expressions

Contact Info

Product

Resources

About