Skip to main content
OSTI.GOVU.S. Department of Energy Office of Scientific and Technical Information
U.S. Department of Energy
Office of Scientific and Technical Information
 

Uncertainty quantification of a deep learning fuel property prediction model

Journal Article · · Applications in Energy and Combustion Science
 [1];  [1];  [1];  [2];  [3];  [3];  [4];  [2];  [3];  [3]
  1. Argonne National Laboratory (ANL), Argonne, IL (United States)
  2. King Abdullah University of Science and Technology (KAUST), Thuwal (Saudi Arabia)
  3. Saudi Arabian Oil Company (Saudi Aramco) (Saudi Arabia)
  4. Aramco Services Co., Novi, MI (United States)
+ Show Author Affiliations
Deep learning models are being widely used in the field of combustion. Given the black-box nature of typical neural network based models, uncertainty quantification (UQ) is critical to ensure the reliability of predictions as well as the training datasets, and for a principled quantification of noise and its various sources. Deep learning surrogate models for predicting properties of chemical compounds and mixtures have been recently shown to be promising for enabling data-driven fuel design and optimization, with the ultimate goal of improving efficiency and lowering emissions from combustion engines. In this study, UQ is performed for a multi-task deep learning model that simultaneously predicts the research octane number (RON), Motor Octane Number (MON), and Yield Sooting Index (YSI) of pure components and multicomponent blends. The deep learning model is comprised of three smaller networks: Extractor 1, Extractor 2, and Predictor, and a mixing operator. The molecular fingerprints of individual components are encoded via Extractor 1 and Extractor 2, the mixing operator generates fingerprints for mixtures/blends based on linear mixing operation, and the predictor maps the fingerprint to the target properties. Two different classes of UQ methods, Monte Carlo ensemble methods and Bayesian neural networks (BNNs), are employed for quantifying the epistemic uncertainty. Combinations of Bernoulli and Gaussian distributions with DropConnect and DropOut techniques are explored as ensemble methods. All the DropConnect, DropOut and Bayesian layers are applied to the predictor network. Aleatoric uncertainty is modeled by assuming that each data point has an independent uncertainty associated with it. The results of the UQ study are further analyzed to compare the performance of BNN and ensemble methods. Although this study is confined to UQ of fuel property prediction, the methodologies are applicable to other deep learning frameworks that are being widely used in the combustion community.
Research Organization:
Argonne National Laboratory (ANL), Argonne, IL (United States)
Sponsoring Organization:
Aramco Americas; Saudi Aramco; USDOE Office of Science
Grant/Contract Number:
AC02-06CH11357
OSTI ID:
2536669
Journal Information:
Applications in Energy and Combustion Science, Journal Name: Applications in Energy and Combustion Science Vol. 16; ISSN 2666-352X
Publisher:
ElsevierCopyright Statement
Country of Publication:
United States
Language:
English

References (27)

Measuring and predicting sooting tendencies of oxygenates, alkanes, alkenes, cycloalkanes, and aromatics on a unified scale journal April 2018
Uncertainty quantification in drug design journal February 2021
Artificial neural network based predictions of cetane number for furanic biofuel additives journal October 2017
Machine learning enabled high-throughput screening of hydrocarbon molecules for the design of next generation fuels journal April 2020
Using machine learning with target-specific feature sets for structure-property relationship modeling of octane numbers and octane sensitivity journal December 2020
Bayesian inference and uncertainty quantification for hydrogen-enriched and lean-premixed combustion systems journal July 2021
Prediction of flammability limit temperatures from molecular structures using a neural network–particle swarm algorithm journal May 2011
Combustion kinetic model uncertainty quantification, propagation and minimization journal April 2015
Combustion machine learning: Principles, progress and prospects journal July 2022
Metrics for Benchmarking and Uncertainty Quantification: Quality, Applicability, and Best Practices for Machine Learning in Chemistry journal February 2021
Identification and description of the uncertainty, variability, bias and influence in quantitative structure-activity relationships (QSARs) for toxicity prediction journal August 2019
Machine Learning and Deep Learning in Chemical Health and Safety: A Systematic Review of Techniques and Applications journal October 2020
Graph Neural Networks for Prediction of Fuel Ignition Quality journal August 2020
Predicting Octane Number Using Nuclear Magnetic Resonance Spectroscopy and Artificial Neural Networks journal April 2018
Artificial Neural Network Based Group Contribution Method for Estimating Cetane and Octane Numbers of Hydrocarbons and Oxygenated Organic Compounds journal October 2017
Data Science Approach to Estimate Enthalpy of Formation of Cyclic Hydrocarbons journal July 2020
Machine Learning To Predict Standard Enthalpy of Formation of Hydrocarbons journal August 2019
Quantitative Structure−Property Relationships for the Prediction of Vapor Pressures of Organic Compounds from Molecular Structures journal May 2000
An Introduction to Variational Methods for Graphical Models journal January 1999
Artificial intelligence-driven design of fuel mixtures journal September 2022
Bayesian semi-supervised learning for uncertainty-calibrated prediction of molecular properties and active learning journal January 2019
Additivity Rules for the Estimation of Molecular Properties. Thermodynamic Properties journal September 1958
Robustness of random forests for regression journal December 2012
On the rational formulation of alternative fuels: melting point and net heat of combustion predictions for fuel compounds using machine learning methods journal April 2013
Hands-On Bayesian Neural Networks—A Tutorial for Deep Learning Users journal May 2022
Could graph neural networks learn better molecular representation for drug discovery? A comparison study of descriptor-based and graph-based models journal February 2021
A hybrid framework for improving uncertainty quantification in deep learning-based QSAR regression modeling journal September 2021

Similar Records

Deeply uncertain: comparing methods of uncertainty quantification in deep learning algorithms
Journal Article · Mon Nov 30 19:00:00 EST 2020 · Machine Learning: Science and Technology · OSTI ID:1731011

Related Subjects

33 ADVANCED PROPULSION SYSTEMS
Aleatoric uncertainty
Bayesian neural network
Deep learning
Epistemic uncertainty
Fuel property prediction
Monte Carlo ensemble methods
Uncertainty quantification