Determination of thermodynamic state variables of liquids from their microscopic structures using an artificial neural network

Que-Salinas, Ulices; Ramírez-González, Pedro E.; Torres-Carbajal, Alexis

doi:10.1039/d0sm02127j

Cited by 7 publications

(3 citation statements)

References 42 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The mean free volume is defined by where the parameter B appears in the GvdW canonical equation of state and is defined as where g ( r ,ρ, T ) is the radial distribution function. Additionally, recent work has shown that Fourier series of RDFs can be used to develop ML models for atomization enthalpy, the prediction of dipole moments, , and the training of ANNs to successfully predict phase behavior of LJ fluids . These examples and many others in the literature present evidence for a strong relationship between self-diffusion and the local structure of a liquid, which can be represented through an RDF.…”

Section: Resultsmentioning

confidence: 99%

Using Computationally-Determined Properties for Machine Learning Prediction of Self-Diffusion Coefficients in Pure Liquids

et al. 2021

View full text Add to dashboard Cite

The ability to predict transport properties of liquids quickly and accurately will greatly improve our understanding of fluid properties both in bulk and complex mixtures, as well as in confined environments. Such information could then be used in the design of materials and processes for applications ranging from energy production and storage to manufacturing processes. As a first step, we consider the use of machine learning (ML) methods to predict the diffusion properties of pure liquids. Recent results have shown that Artificial Neural Networks (ANNs) can effectively predict the diffusion of pure compounds based on the use of experimental properties as the model inputs. In the current study, a similar ANN approach is applied to modeling diffusion of pure liquids using fluid properties obtained exclusively from molecular simulations. A diverse set of 102 pure liquids is considered, ranging from small polar molecules (e.g., water) to large nonpolar molecules (e.g., octane). Self-diffusion coefficients were obtained from classical molecular dynamics (MD) simulations. Since nearly all the molecules are organic compounds, a general set of force field parameters for organic molecules was used. The MD methods are validated by comparing physical and thermodynamic properties with experiment. Computational input features for the ANN include physical properties obtained from the MD simulations as well as molecular properties from quantum calculations of individual molecules. Fluid properties describing the local liquid structure were obtained from center of mass radial distribution functions (COM-RDFs). Feature sensitivity analysis revealed that isothermal compressibility, heat of vaporization, and the thermal expansion coefficient were the most impactful properties used as input for the ANN model to predict the MD simulated self-diffusion coefficients. The MD-based ANN successfully predicts the MD self-diffusion coefficients with only a subset (2 to 3) of the available computationally determined input features required. A separate ANN model was developed using literature experimental self-diffusion coefficients as model targets. Although this second ML model was not as successful due to a limited number of data points, a good correlation is still observed between experimental and ML predicted self-diffusion coefficients.

show abstract

Section: Resultsmentioning

confidence: 99%

Using Computationally-Determined Properties for Machine Learning Prediction of Self-Diffusion Coefficients in Pure Liquids

et al. 2021

View full text Add to dashboard Cite

show abstract

“…Therefore, we focus on the LJ fluid to make a database to train the ANN. The database is composed of several g(r) determined at different thermodynamic states that spans a wide region in the phase diagram [34].…”

Section: Molecular Model Of Simple Liquids and Training Data Setmentioning

confidence: 99%

Prediction of equations of state of molecular liquids by an artificial neural network.

2022

View full text Add to dashboard Cite

In this work an artificial neural network (ANN) was used to determine the pressure and internal energy equations of state of noble gases and some molecular liquids by predicting thermodynamic state variables like density and temperature encoded in the radial distribution function. The ANN is trained to predict the thermodynamic state variables using only the structural data. Then, predicted values are used to compute equations of state of real liquids such as argon, neon, krypton and xenon as well as some molecular liquids like nitrogen, carbon dioxide, methane and ethylene. In order to assess the ANN predictions the relative percentage error with the exact values were determined, showing that its magnitude is less than 1%. Thus, the comparison between equations of state computed with the predicted variables and experimental results exhibits a very good agreement for most of the liquids studied here. Since our ANN implementation only requires the microscopic structure as an input, data incoming from experiments, theoretical frameworks or simulations are suitable to perform predictions of state variables and with that complement the thermodynamic characterisation of liquids through the determination of equations of state. Moreover, further improvements or extensions related with the microscopic structure database can be safely addressed without changing the neural network architecture presented here.

show abstract

“…[39,43] Thus, graph network can replace the existing molecular dynamics derivation process to study cluster dynamics specific processes through predictive methods. [44,45] This study aims to apply a novel machine learning approach to a classical physics problem. [46] We hope to obtain the final state of cluster motion at a specified time step using the predictive capability of graph network models with only the input of initial bit patterns.…”

Section: Introductionmentioning

confidence: 99%

Graph dynamical networks for forecasting collective behavior of active matter

Liu¹,

Wang²,

Zhao³

et al. 2022

Chinese Phys. B

View full text Add to dashboard Cite

After decades of theoretical studies, the rich phase states of active matter and cluster kinetic processes are still of research interest. And how to efficiently calculate the dynamical processes under their complex conditions becomes an open problem. Recently, machine learning methods have been proposed to predict the degree of coherence of active matter systems. In this way, the phase transition process of the system is quantified and studied. In this paper, we use Graph Network as a powerful model to determine the evolution of active matter with variable individual velocities solely based on the initial position and state of the particles. The Graph Network accurately predicts the order parameters of the system in different scale models with different individual velocities, noise and density to effectively evaluate the effect of diverse condition. Compared with the classical physical deduction method, we demonstrate that Graph Network prediction is excellent, which could save significantly computing resources and time. In addition to active matter, our method can be applied widely to other large-scale physical systems.

show abstract

Determination of thermodynamic state variables of liquids from their microscopic structures using an artificial neural network

Abstract: In this work we implement a machine learning method based on an artificial neural network to predict the thermodynamic state of a liquid using only its microscopic structure provided by the radial distribution function.

Cited by 7 publications

References 42 publications

Using Computationally-Determined Properties for Machine Learning Prediction of Self-Diffusion Coefficients in Pure Liquids

Using Computationally-Determined Properties for Machine Learning Prediction of Self-Diffusion Coefficients in Pure Liquids

Prediction of equations of state of molecular liquids by an artificial neural network.

Graph dynamical networks for forecasting collective behavior of active matter

Contact Info

Product

Resources

About