A Methodology to Diagnose Process/Model Mismatch in First-Principles Models

Meneghetti, Natascia; Facco, Pierantonio; Bezzo, Fabrizio; Barolo, Massimiliano

doi:10.1021/ie501812c

Cited by 11 publications

(6 citation statements)

References 23 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the case of knowledge-driven models, could a data-driven approach be used to model these less-well modeled parts? Another issue regards “model diagnosis”, that is, finding the root cause of process-model mismatch that may arise when a knowledge-driven model is challenged by a set of experimental data Are there sufficient measurements?…”

Section: Open Research Problemsmentioning

confidence: 99%

Linking Models and Experiments

Bonvin

Georgakis

Pantelides

et al. 2016

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

This position paper gives an overview of the discussion that took place at FIPSE 2 at Aldemar Resort, east of Heraklion, Crete, in June 21–23, 2014. This is the second conference in the series “Future Innovation in Process Systems Engineering” (http://fi-in-pse.org), which takes place every other year in Greece, with the objective to discuss open research challenges in three topics in Process Systems Engineering. One of the topics of FIPSE 2 was the issue of “Linking Models and Experiments”, which is described in this publication. Process models have been used extensively in academia and industry for several decades. Yet, this paper argues that there are still substantial challenges to be addressed along the lines of model structure selection, identifiability, experiment design, nonlinear parameter estimation, model validation, model improvement, online model adaptation, model portability, modeling of complex systems, numerical methods, software environments, and implementation aspects. Although there has been an exponential increase in the number of publications dealing with “modeling”, the majority of these publications do not use sound statistical tools to evaluate the model quality and accuracy and also present modeling as a noniterative task. As a result, the models often have either too few or too many parameters, thus requiring trimming down or enhancing before they can be used appropriately. Also, this position paper argues that the models should be developed with a purpose in mind, as, for example, different models are needed for design, control, monitoring, and optimization

show abstract

Section: Open Research Problemsmentioning

confidence: 99%

Linking Models and Experiments

Bonvin

Georgakis

Pantelides

et al. 2016

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

show abstract

“…Process modeling as a method to translate process knowledge into appropriate mathematical representation is an important tool to support many process engineering activities like process design, simulation, monitoring, and control 1,2 . The majority of chemical processes are complex in nature because a large number of reactions, nonlinear equations, implicit correlations, and variables are involved 3 .…”

Section: Introductionmentioning

confidence: 99%

“…With an increasingly large number of WB models being developed and implemented, plant‐model mismatch (PMM) has become a challenge. When the model prediction or output of a particular process is compared to historical plant data, they may not agree with each other, resulting in a PMM 1 . PMM identification has been used in various application areas such as Model Predictive Control (MPC), 20 real‐time optimization, 21 uncertainty quantification, 22 and state estimation 23‐25 .…”

Section: Introductionmentioning

confidence: 99%

“…PMM identification has been used in various application areas such as Model Predictive Control (MPC), 20 real‐time optimization, 21 uncertainty quantification, 22 and state estimation 23‐25 . The underlying reasons for PMM may be attributed to limited process knowledge and inappropriate development of models (including but not limited to oversimplification, wrongly identified parameters, mismatch in physical properties) 1 . The importance of identifying PMM has been recognized in many process control and reaction system applications, 26,27 but it has not been applied to guide the development of hybrid models.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A framework of hybrid model development with identification of plant‐model mismatch

Chen

Ierapetritou

2020

AIChE Journal

View full text Add to dashboard Cite

Hybrid modeling has attracted increasing attention in order to take advantage of the additional data to improve process understanding. Current practice often adopts mechanistic models to predict process behaviors. These mechanistic models are based on physical understandings and experimental studies, but they sometimes lead to plant‐model mismatch (PMM) as they may be inaccurate to fully describe real processes. Black‐box models can serve as an alternative, but they often suffer from poor generalization and interpretability. To combine the two techniques, hybrid models are developed to make use of process data while maintaining a degree of physical understanding. In this work, we implement a framework of identification of PMM using partial correlation coefficient and mutual information, followed by introducing and comparing serial, parallel, and combined structures of hybrid models. The framework is applied and tested with a simulated reactor model and two pharmaceutical unit operation case studies.

show abstract

“…The importance of diagnosing process-model mismatch is recognised in the literature on process monitoring (Badwe et al, 2009;Wang et al, 2012), but the problem is considered only in the context of linear, black-box models for control applications. An approach for diagnosing process-model mismatch in phenomenological models was proposed by Meneghetti et al (2014), where a latent variable model is used to detect differences between process and model in the distribution of some auxiliary variables. These auxiliary variables represent Figure 1: Tuning of complexity in parametric models.…”

Section: Introductionmentioning

confidence: 99%

A diagnostic procedure for improving the structure of approximated kinetic models

Quaglio

Fraga

Galvanin

2020

Computers & Chemical Engineering

View full text Add to dashboard Cite

Kinetic models of chemical and biochemical phenomena are frequently built from simplifying assumptions. Whenever a model is falsified by data, its mathematical structure should be modified embracing the available experimental evidence. A framework based on maximum likelihood inference is illustrated in this work for diagnosing model misspecification and improving the structure of approximated models. In the proposed framework, statistical evidence provides a measure to justify a modification of the model structure, namely a reduction of complexity through the removal of irrelevant parameters and/or an increase of complexity through the replacement of relevant parameters with more complex state-dependent expressions. A tailored Lagrange multipliers test is proposed to support the scientist in the improvement of parametric models when an increase in model complexity is required.

show abstract

A Methodology to Diagnose Process/Model Mismatch in First-Principles Models

Cited by 11 publications

References 23 publications

Linking Models and Experiments

Linking Models and Experiments

A framework of hybrid model development with identification of plant‐model mismatch

A diagnostic procedure for improving the structure of approximated kinetic models

Contact Info

Product

Resources

About