Data-driven surrogate modeling and benchmarking for process equipment

Gonçalves, G.F.N.; Batchvarov, Assen; Liu, Yuyi; Liu, Yuxin; Mason, Lachlan; Pan, Indranil; Matar, Omar

doi:10.1017/dce.2020.8

Cited by 8 publications

(3 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We would suggest these databases should be gradually improved and become more and more feasible for all users in the future. At the moment, one may improve the training performance by choosing a suitable data sampling/handling strategy specified to their data characteristics/qualities/distributions, such as random sampling, greedy sampling on (both) inputs/output, and variational sampling . The last option is that the theory-based methods can provide large data sets for training ML, which has become an effective methodology for calculations of rate parameters of kinetic models.…”

Section: Current Status and Challengesmentioning

confidence: 99%

Review of Machine Learning for Hydrodynamics, Transport, and Reactions in Multiphase Flows and Reactors

Zhu

Chen

Ouyang

et al. 2022

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

Artificial intelligence (AI), machine learning (ML), and data science are leading to a promising transformative paradigm. ML, especially deep learning and physics-informed ML, is a valuable toolkit that complements incomplete domain-specific knowledge in conventional experimental and computational methods. ML can provide flexible techniques to facilitate the conceptual development of new robust predictive models for multiphase flows and reactors by finding hidden pattern/information/mechanism in a data set. Due to such emergence, we thereby comprehensively survey, explore, analyze, and discuss key advancements of recent ML applications to hydrodynamics, heat and mass transfer, and reactions in single-phase and multiphase flow systems from different aspects: (1) development of multiphase closure models of drag force, turbulence stresses and heat/mass transfer to improve the accuracy and efficiency of typical CFD simulations; (2) image reconstruction, regime identification, key parameter predictions, and optimization of multiphase flow and transport fields; (3) reaction kinetics modeling (e.g., predictions of reaction networks, kinetic parameters, and species production) and reaction condition optimization. These sections also discuss and analyze the key advantages and weakness of ML for solving the problems in the domain of multiphase flows and reactors. Finally, we summarize the under-solving challenges and opportunities in order to identify future directions that would be useful for the research community. Future development and study of multiphase flows and reactors are envisaged to be accelerated by ML and data science.

show abstract

Section: Current Status and Challengesmentioning

confidence: 99%

Review of Machine Learning for Hydrodynamics, Transport, and Reactions in Multiphase Flows and Reactors

Zhu

Chen

Ouyang

et al. 2022

Ind. Eng. Chem. Res.

View full text Add to dashboard Cite

show abstract

“…Owoyele et al 23 used AL to perform simulation‐based data generation, ML learning and surrogate optimization to refine solution in the vicinity of predicted optimum parameters for design of a compression ignition engine. Gonçalves et al 24 studied the generation of simulation‐based surrogate models with the task of parameter domain exploration using various sampling and regression‐based AL strategies. In a similar study, Pan et al 25 used AL for developing surrogate models for industrial fluid flow case studies under a constraint of a limited function evaluations.…”

Section: Comparison To Related Workmentioning

confidence: 99%

Iterative symbolic regression for learning transport equations

et al. 2022

View full text Add to dashboard Cite

Computational fluid dynamics (CFD) analysis is widely used in chemical engineering. Although CFD calculations are accurate, the computational cost associated with complex systems makes it difficult to obtain empirical equations between system variables. Here, we combine active learning (AL) and symbolic regression (SR) to get a symbolic equation for system variables from CFD simulations. Gaussian process regression‐based AL allows for automated selection of variables by selecting the most instructive points from the available range of possible parameters. The results from these experiments are then passed to SR to find empirical symbolic equations for CFD models. This approach is scalable and applicable for any desired number of CFD design parameters. To demonstrate the effectiveness, we use this method with two model systems. We recover an empirical equation for the pressure drop in a bent pipe and a new equation for predicting backflow in a heart valve under aortic insufficiency.

show abstract

“…These new variables have a reduced dimension, containing most of the information in the old variables. The old and new variables are independent of each other [17]. The specific operation process is shown below.…”

Section: Principal Component Analysis (Pca)mentioning

confidence: 99%

Construction of an Electromechanical Automation and Fault Diagnosis System Using Data-Driven Technology in the Context of Big Data

2021

IJOMAM

View full text Add to dashboard Cite

The sensor faults in electromechanical equipment are diagnosed and predicted to accelerate the industrial automation and stable operation of electromechanical equipment. Firstly, several basic fault prediction methods are introduced, and the data of electromechanical equipment during production can be well diagnosed and processed through data-driven technology. Secondly, the principal component analysis (PCA) method is introduced, and accordingly, the optimized dynamic principal component analysis (DPCA) method and several common sensor faults are proposed. Thirdly, the specific steps of the PCA and DPCA methods are discussed in sensor faults diagnosis, respectively. The PCA method can study the multivariate nonlinear data concurrently. On this basis, the optimized DPCA method can analyse the dynamic features of the real-time data collected by sensors. Finally, the fault type that produces a fixed deviation is studied. Then, the experiment of the draw-wire displacement sensor is simulated through the optimized PCA and DPCA. The square prediction error (SPE) and prediction-error rate are compared with the error. The results show that both methods can predict the fault of the sensor. The results of the T 2 statistics obtained through the PCA show that the sensor fault can be effectively predicted, but the prediction-error rate and fault-omission rate are also high. In the DPCA algorithm, the SPE statistics show that the DPCA algorithm can predict the fault with a small prediction-error rate and faultomission rate, which is far lower than the PCA algorithm. Therefore, dynamic factors analysis can improve sensor fault prediction, which is of great significance to the acceleration of industrial automation and equipment intelligence.

show abstract

Data-driven surrogate modeling and benchmarking for process equipment

Cited by 8 publications

References 15 publications

Review of Machine Learning for Hydrodynamics, Transport, and Reactions in Multiphase Flows and Reactors

Review of Machine Learning for Hydrodynamics, Transport, and Reactions in Multiphase Flows and Reactors

Iterative symbolic regression for learning transport equations

Construction of an Electromechanical Automation and Fault Diagnosis System Using Data-Driven Technology in the Context of Big Data

Contact Info

Product

Resources

About