Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) Method
In this work, a data-driven methodology for modeling combustion kinetics, Learned Intelligent Tabulation (LIT), is presented. LIT aims to accelerate the tabulation of combustion mechanisms via machine learning algorithms such as Deep Neural Networks (DNNs). The high-dimensional composition space is...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2021-11-01
|
| Series: | Energies |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1996-1073/14/23/7851 |
| _version_ | 1797507916719194112 |
|---|---|
| author | Majid Haghshenas Peetak Mitra Niccolò Dal Santo David P. Schmidt |
| author_facet | Majid Haghshenas Peetak Mitra Niccolò Dal Santo David P. Schmidt |
| author_sort | Majid Haghshenas |
| collection | DOAJ |
| description | In this work, a data-driven methodology for modeling combustion kinetics, Learned Intelligent Tabulation (LIT), is presented. LIT aims to accelerate the tabulation of combustion mechanisms via machine learning algorithms such as Deep Neural Networks (DNNs). The high-dimensional composition space is sampled from high-fidelity simulations covering a wide range of initial conditions to train these DNNs. The input data are clustered into subspaces, while each subspace is trained with a DNN regression model targeted to a particular part of the high-dimensional composition space. This localized approach has proven to be more tractable than having a global ANN regression model, which fails to generalize across various composition spaces. The clustering is performed using an unsupervised method, Self-Organizing Map (SOM), which automatically subdivides the space. A dense network comprised of fully connected layers is considered for the regression model, while the network hyper parameters are optimized using Bayesian optimization. A nonlinear transformation of the parameters is used to improve sensitivity to minor species and enhance the prediction of ignition delay. The LIT method is employed to model the chemistry kinetics of zero-dimensional <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msub><mi mathvariant="normal">H</mi><mn>2</mn></msub><mo>–</mo><msub><mi mathvariant="normal">O</mi><mn>2</mn></msub></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>CH</mi><mn>4</mn></msub></semantics></math></inline-formula>-air combustion. The data-driven method achieves good agreement with the benchmark method while being cheaper in terms of computational cost. LIT is naturally extensible to different combustion models such as flamelet and PDF transport models. |
| first_indexed | 2024-03-10T04:55:15Z |
| format | Article |
| id | doaj.art-64f01476a60e4dc888e69c168b257b6d |
| institution | Directory Open Access Journal |
| issn | 1996-1073 |
| language | English |
| last_indexed | 2024-03-10T04:55:15Z |
| publishDate | 2021-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Energies |
| spelling | doaj.art-64f01476a60e4dc888e69c168b257b6d2023-11-23T02:18:43ZengMDPI AGEnergies1996-10732021-11-011423785110.3390/en14237851Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) MethodMajid Haghshenas0Peetak Mitra1Niccolò Dal Santo2David P. Schmidt3Mechanical and Industrial Engineering Department, University of Massachusetts Amherst, Amherst, MA 01003, USAMechanical and Industrial Engineering Department, University of Massachusetts Amherst, Amherst, MA 01003, USAMathWorks Inc., Cambridge CB4 0HH, UKMechanical and Industrial Engineering Department, University of Massachusetts Amherst, Amherst, MA 01003, USAIn this work, a data-driven methodology for modeling combustion kinetics, Learned Intelligent Tabulation (LIT), is presented. LIT aims to accelerate the tabulation of combustion mechanisms via machine learning algorithms such as Deep Neural Networks (DNNs). The high-dimensional composition space is sampled from high-fidelity simulations covering a wide range of initial conditions to train these DNNs. The input data are clustered into subspaces, while each subspace is trained with a DNN regression model targeted to a particular part of the high-dimensional composition space. This localized approach has proven to be more tractable than having a global ANN regression model, which fails to generalize across various composition spaces. The clustering is performed using an unsupervised method, Self-Organizing Map (SOM), which automatically subdivides the space. A dense network comprised of fully connected layers is considered for the regression model, while the network hyper parameters are optimized using Bayesian optimization. A nonlinear transformation of the parameters is used to improve sensitivity to minor species and enhance the prediction of ignition delay. The LIT method is employed to model the chemistry kinetics of zero-dimensional <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><mrow><msub><mi mathvariant="normal">H</mi><mn>2</mn></msub><mo>–</mo><msub><mi mathvariant="normal">O</mi><mn>2</mn></msub></mrow></semantics></math></inline-formula> and <inline-formula><math xmlns="http://www.w3.org/1998/Math/MathML" display="inline"><semantics><msub><mi>CH</mi><mn>4</mn></msub></semantics></math></inline-formula>-air combustion. The data-driven method achieves good agreement with the benchmark method while being cheaper in terms of computational cost. LIT is naturally extensible to different combustion models such as flamelet and PDF transport models.https://www.mdpi.com/1996-1073/14/23/7851combustionkineticsmachine learningneural network (NN)CFD |
| spellingShingle | Majid Haghshenas Peetak Mitra Niccolò Dal Santo David P. Schmidt Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) Method Energies combustion kinetics machine learning neural network (NN) CFD |
| title | Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) Method |
| title_full | Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) Method |
| title_fullStr | Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) Method |
| title_full_unstemmed | Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) Method |
| title_short | Acceleration of Chemical Kinetics Computation with the Learned Intelligent Tabulation (LIT) Method |
| title_sort | acceleration of chemical kinetics computation with the learned intelligent tabulation lit method |
| topic | combustion kinetics machine learning neural network (NN) CFD |
| url | https://www.mdpi.com/1996-1073/14/23/7851 |
| work_keys_str_mv | AT majidhaghshenas accelerationofchemicalkineticscomputationwiththelearnedintelligenttabulationlitmethod AT peetakmitra accelerationofchemicalkineticscomputationwiththelearnedintelligenttabulationlitmethod AT niccolodalsanto accelerationofchemicalkineticscomputationwiththelearnedintelligenttabulationlitmethod AT davidpschmidt accelerationofchemicalkineticscomputationwiththelearnedintelligenttabulationlitmethod |