Data-Driven Dynamic Modeling of Batch Processes Having Different Initial Conditions and Missing Measurements

Shokry, Ahmed; Pérez-Moya, Montserrat; Graells, Moisès; Espuña, Antonio

doi:10.1016/b978-0-444-63965-3.50074-x

Cited by 6 publications

(7 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The method was successfully applied to model different batch processes for the purpose of dynamic optimization [51] and continuous processes [52] for the purpose of Fault Detection and Diagnosis (FDD). Shokry et al [53] successfully applied the same methodology to the univariate dynamic modelling of a real batch process operated under different initial conditions and involving missing measurements. The obtained model is used as a dynamic observer for the online supervision of the process and the detection of possible faults.…”

Section: Review On Data-driven Dynamic Modelling In Chemical Processesmentioning

confidence: 99%

“…For real situation, where a database of the process variables measurements history is available, the training data selection should cover as much as possible the dynamic conditions of the process, in order to feed model with sufficient information about the process [53]. Finally, it should be mentioned that in all the analyzed cases but, especially, in situations where only few input-output signals are available for the training and/or they may represent a biased or partial view of the process (as in the last case study, see Figure 16-(e,f), red crosses), the resulting dynamic models may be very sensible to the eventual evolution of the real process behavior through the time, which may drive it to unexpected/unexplored conditions, either due to the natural evolution of the process (e.g.…”

Section: Oil Shale Pyrolysismentioning

confidence: 99%

“…1,38 For real situations where a database of the process variables' measurement history is available, the training data selection should cover as much as possible the dynamic conditions of the process in order to feed the model with sufficient information about the process. 53 Figure 17a shows the computational time required for the training data generation using the proposed DOCE procedure. The time required for generating the other type of training data (input−output signal, see Figure 13) is (again) constant and equals to an average of 9.0 s. Figure 17b illustrates the escalation of the training time with the increase of the dynamic models' lag and that the training times of the OK-based dynamic models (mauve color) are (again) larger relative to the ANN (green color) escalation.…”

Section: Applicationsmentioning

confidence: 99%

See 2 more Smart Citations

Dynamic Surrogate Modeling for Multistep-ahead Prediction of Multivariate Nonlinear Chemical Processes

Shokry

Baraldi

Zio

et al. 2020

Ind. Eng. Chem. Res.

Self Cite

View full text Add to dashboard Cite

This work proposes a methodology for multivariate dynamic modeling and multistep-ahead prediction of nonlinear systems using surrogate models for the application to nonlinear chemical processes. The methodology provides a systematic and robust procedure for the development of data-driven dynamic models capable of predicting the process outputs over long time horizons. It is based on using surrogate models to construct several Nonlinear AutoRegressive eXogenous models (NARX), each one approximating the future behavior of one process output as a function of the current and previous process inputs and outputs. The developed dynamic models are employed in a recursive schema to predict the process future outputs over several time steps (multistep-ahead prediction). The methodology is able to manage two different scenarios: 1) one in which a set of input-output signals collected from the process is only available for training, and 2) another in which a mathematical model of the process is available and can be used to generate specific datasets for training. With respect to the latter, the proposed methodology includes a specific procedure for the selection of training data in dynamic modeling based on Design Of Computer Experiment (DOCE) techniques. The proposed methodology is applied to case studies from the process industry presented in the literature. The results show very high prediction accuracies over long time horizons. Also, thanks to the flexibility, robustness and computational efficiency of surrogate modeling, the methodology allows dealing with a wide range of situations, which would be difficult to address using first principle models.

show abstract

Section: Review On Data-driven Dynamic Modelling In Chemical Processesmentioning

confidence: 99%

Section: Oil Shale Pyrolysismentioning

confidence: 99%

Section: Applicationsmentioning

confidence: 99%

See 1 more Smart Citation

Dynamic Surrogate Modeling for Multistep-ahead Prediction of Multivariate Nonlinear Chemical Processes

Shokry

Baraldi

Zio

et al. 2020

Ind. Eng. Chem. Res.

Self Cite

View full text Add to dashboard Cite

show abstract

“…More recently, semi-empirical models combining a first principles modeling approach with empirical observations have been proposed by Cabrera Reina et al (2012Reina et al ( , 2015. Finally, advanced multivariate techniques named data-based modeling (DBM) can also be found (Shokry et al 2015(Shokry et al , 2017.…”

Section: Introductionmentioning

confidence: 99%

A novel modeling approach for a generalizable photo-Fenton-based degradation of organic compounds

Audino

Pérez-Moya

Graells

et al. 2020

Environ Sci Pollut Res

Self Cite

View full text Add to dashboard Cite

This work aims at proposing and validating a model that can be exploited for the future development of industrial applications (e.g., process design and control) of Fenton and photo-Fenton processes. Hence, a compromise modeling solution has been developed between the non-generalizable accuracy of the first principles models (FPMs) and the oversimplification of the empirical models (EMs). The work presents a novel model of moderate complexity that is simplified enough to be generalizable and computationally affordable, while retaining physical meaning. The methodology is based on a general degradation mechanism that can be algorithmically generated from the carbon number of the target compound, as well as from the knowledge of two kinetic parameters, one for the faster initial rate and the other one for the subsequent degradation steps. The contaminant degradation mechanism has been combined with an appropriately simplified implementation of the well-known Fenton and photo-Fenton kinetics. This model describes the degradation not only of the target compound and of the oxidant, but also of total organic carbon (TOC), which is used to define the overall quality of the water. Experimental design techniques were used along with a non-conventional modeling methodology of programmable process structures (PPS). This novel modeling approach was applied and validated on the degradation of three model compounds. A successful prediction of the evolution of the contaminants H 2 O 2 and TOC was confirmed and assessed by the root mean square error (RMSE).

show abstract

“…: Kaneko and Funatsu, 2013, to update/improve the soft-sensor) or even online (e.g. : Shokry et al, 2017b, to confirm estimations and eventually reset them to better match the current situation).…”

Section: Introductionmentioning

confidence: 99%

Data-driven soft-sensors for online monitoring of batch processes with different initial conditions

Shokry

Vicente

Escudero

et al. 2018

Computers & Chemical Engineering

Self Cite

View full text Add to dashboard Cite

A soft-sensing methodology applicable to batch processes operated under changeable initial conditions is presented. These cases appear when the raw materials specifications differ from batch to batch, different production scenarios should be managed, etc. The proposal exploits the capabilities of the machine learning techniques to provide practical soft-sensing approach with minimum tuning effort in spite of the fact that the inherent dynamic behavior of batch systems are tracked through other online indirect measurements. Current data modelling techniques have been also tested within the proposed methodology to demonstrate its advantages. Two simulation casestudies and a pilot-plant case-study involving a complex batch process for wastewater treatment are used to illustrate the problem, to assess the modelling approach and to compare the modelling techniques. The results reflect a promising accuracy even when the training information is scarce, allowing significant reductions in the cost associated to batch processes monitoring and control.

show abstract

Data-Driven Dynamic Modeling of Batch Processes Having Different Initial Conditions and Missing Measurements

Cited by 6 publications

References 5 publications

Dynamic Surrogate Modeling for Multistep-ahead Prediction of Multivariate Nonlinear Chemical Processes

Dynamic Surrogate Modeling for Multistep-ahead Prediction of Multivariate Nonlinear Chemical Processes

A novel modeling approach for a generalizable photo-Fenton-based degradation of organic compounds

Data-driven soft-sensors for online monitoring of batch processes with different initial conditions

Contact Info

Product

Resources

About