Semi-mechanistic modeling of chemical processes with neural networks

Braake, H.A.B. te; Can, H.J.L. van; Verbruggen, H.B.

doi:10.1016/s0952-1976(98)00011-6

Cited by 51 publications

(26 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The probably most efficient way is to aim from the beginning at the development of a hybrid semi-parametric model, because it can be tented to develop the model in such a way that good inter-and extrapolation properties are yielded with a small number of specifically designed experiments at both scales, e.g. Braake et al (1998). Another strategy (Bollas et al, 2003;Simon et al, 2006) is to complement a mechanistic model, developed at small-scale, with nonparametric techniques that represent specific parts of the model on the large-scale.…”

Section: Scale-upmentioning

confidence: 99%

“…white-box models) into empirical (black-box) models (Bohlin & Graebe, 1995;Jorgensen & Hangos, 1995;Tulleken, 1993). According to Braake, van Can, and Verbruggen (1998) a grey-box model is based on the same unstructured nature as a black-box model. The term has however evolved to designate all types of models that combine white-box and black-box submodels.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Hybrid semi-parametric modeling in process systems engineering: Past, present and future

Stosch¹,

Oliveira²,

Peres³

et al. 2014

Computers & Chemical Engineering

360

201

View full text Add to dashboard Cite

Hybrid semi-parametric models consist of model structures that combine parametric and nonparametric submodels based on different knowledge sources. The development of a hybrid semi-parametric model can offer several advantages over traditional mechanistic or data-driven modeling, as reviewed in this paper. These advantages, such as broader knowledge base, transparency of the modeling approach and cost-effective model development, have been widely recognized, not only in academia but also in the industry.In this paper, the most common hybrid semi-parametric modeling and parameter identification techniques are revisited. Applications in the areas of (bio)chemical engineering for process monitoring, control, optimization, scaleup and model-reduction are reviewed. It is outlined that the application of hybrid semi-parametric techniques does not automatically lead into better results but that rational knowledge integration has potential to significantly improve model-based process operation and design.

show abstract

Section: Scale-upmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Hybrid semi-parametric modeling in process systems engineering: Past, present and future

Stosch¹,

Oliveira²,

Peres³

et al. 2014

Computers & Chemical Engineering

360

201

View full text Add to dashboard Cite

show abstract

“…Because of an empirical model's best-fit nature, in general it cannot be used to predict the outcome of an experiment done outside of the sample data used to fit the model. Empirical and mechanistic models are not always developed separately, such as with semimechanistic 3 and grey-box models. 4 These two studies are focused on developing models from existing data, while very little effort has been given to experimental design to improve the models.…”

Section: Introductionmentioning

confidence: 99%

A new approach to batch process optimization using experimental design

Wissmann

Grover

2008

AIChE Journal

View full text Add to dashboard Cite

Empirical and mechanistic experimental design methods are combined to construct partial models, which are, thus, used to design a process. The grid algorithm restricts the next experimental point to potential process optima, according to the confidence intervals around the optimal points, and works with any experimental design algorithm such as D‐optimal. Two case studies show the advantages of implementing the grid algorithm. On average the improvement due to the grid algorithm was 15–20% in the first case study. The second case study is based on thin film growth using four potential models, with the most probable model used for experimental design. The grid algorithm balances the trade‐off between two extremes: D‐optimal designs and sampling at the predicted optimal point. The methodology presented shows that the experimenter does not have to decide ahead of time on purely empirical or mechanistic experimental design methods, since both may be useful. © 2008 American Institute of Chemical Engineers AIChE J, 2009

show abstract

“…In real plant optimization, there is a clear need for combining known mechanistic models together with available operating data, in a way that goes further than just doing parameter adjustment in the mechanistic model. But when using a parallel or serial combination of a mechanistic model and an artificial neural network (or other empirical based modules) the resulting model behavior in extrapolation is known to be unstable and non-reliable (Braake, van Can, & Verbruggen, 1998). With the internal simple empirical extensions that we will propose in our framework, to be described in this paper, it is possible for one to keep the main relations between variables, provided by an initial mechanistic model of the plant, but obtain from it a derived model with a better data fitting and prediction capabilities, and therefore able to provide a closer representation of a real industrial plant.…”

Section: Introductionmentioning

confidence: 99%

A semi-mechanistic model building framework based on selective and localized model extensions

Lima

Saraiva

2007

Computers & Chemical Engineering

View full text Add to dashboard Cite

In the core of many process systems engineering tasks, like design, control, optimization and fault diagnosis, a mathematical model of the underlying plant plays a key role. Such models are so important that extensive studies are available, recommending different modeling techniques to be adopted for specific processes or goals. It is usual and practical to split modeling techniques under two main groups: mechanistic methods and empirical or statistical methods. Both paradigms have been adopted, but very few frameworks were developed to combine and integrate features from both of them. In this article we describe a framework for data-driven evolution of static mechanistic models with a selective inclusion of simple empirical terms. To illustrate its practical potential, our framework is applied to the identification of a non-ideal reactor and to the optimization of the Otto-Williams benchmark reactor.

show abstract

Semi-mechanistic modeling of chemical processes with neural networks

Cited by 51 publications

References 9 publications

Hybrid semi-parametric modeling in process systems engineering: Past, present and future

Hybrid semi-parametric modeling in process systems engineering: Past, present and future

A new approach to batch process optimization using experimental design

A semi-mechanistic model building framework based on selective and localized model extensions

Contact Info

Product

Resources

About