Global ETD Search
Search theses and dissertations gathered from participating repositories worldwide. Every result links back to the library that holds it. No account is needed.
Results
Showing 1 to 17 of 17 for “"Neural network potentials"”.
-
Neural Network Potentials for Molecule- Surface Interactions
… framework TensorFlow was used to generate neural network representations of the energy surface from this data set. The final representation accurately reproduces the features of the underlying PES around potential entry channels for hydrogen dissociation, which agrees well with other …
-
Neural Network Potentials for Atomistic Simulations of Reactive Chemistry
… for training datasets used to train ML potentials, specifically to calculate chemical kinetics of the OH+ CH4 hydrogen abstraction reaction. In addition, combined ML and molecular mechanics methods for condensed phase reactions is discussed.
-
Development and Assessment of Interatomic Neural Network Potentials for Reactive Chemistry and Molecular Dynamics Simulations
… The development of interatomic machine learning potentials has resulted in the ability to model high-quality potential energy surfaces with near ab initio level of accuracy at low computational cost. However, just like other machine learning applications, such methods face challenges when it …
-
Probing Ion Effects In Nanoconfined Aqueous Electrolytes: A Molecular Dynamics Study Using Neural Network Potentials
… we underscore the importance of machine learning potentials in accurately predicting the complex behaviors exhibited by such confined systems, which the classical force fields might overlook.
-
Characterizing the potential energy surface of two dimensional and bulk materials using high dimensional neural network potentials
… is obtained using a feed-forward artificial neural network. An approximate Density Functional Theory method, Density Functional Tight Binding (DFTB+), is used to compute quantities required for the reference dataset. It is found that a network made up of linear activation functions in ænet is …
-
Data-driven analysis and atomistic simulations of next-generation materials for energy conversion and storage
… of dynamic studies, the machine learning potentials (MLPs), namely high-dimensional committee neural network potentials (CNNPs) were developed for 2D MOFs for long-time and large-scale MD simulations. Active learning was used to maintain accuracy while reducing training costs. The CNNP …
-
A Neural Network Potential Capable of Predicting Short-Range Interactions and Long-Range Dispersion Interactions Involving Organic Halides and Sulfides
Neural Network Potentials (NNPs) have recently emerged as useful computational methods for predicting intermolecular interactions. By predicting the intermolecular interactions directly, they can perform calculations on chemical systems at a significantly lower computational cost compared to …
-
Discovering Ice and Water Structures with High-Resolution AFM, Atomistic Modeling, and Machine Learning
… and Cu(111) surfaces, centered about the use of neural network potentials. Then, molecular dynamics simulations are carried out to uncover the atomistic structure of cellulose-Iα and α-chitin nanocrystals surfaces in water. Finally, a high-throughput workflow is developed to identify the …
-
TrIP - Transformer Interatomic Potential Predicts Realistic Energy Surface Using Physical Bias
… algorithms holds promise for producing general potentials that support dozens of atomic species. I present my own Transformer Interatomic Potential (TrIP): a chemically sound potential based on the SE(3)-Transformer. TrIP's species-agnostic architecture--using continuous atomic representation …
-
Intermolecular Interactions in Transferable Neural Network Potential
Transferable neural network potentials (NNPs) offer a promising alternative to traditional quantum mechanical calculations by delivering comparable accuracy at significantly reduced computational cost. However, earlier generations of NNPs are constrained by their architectures and training …
-
Advancing on-the-fly Machine Learned Force Fields and Leveraging Electronic Structure Theory for Computational Catalysis
… this space, and existing machine learned potentials are not sufficiently general to provide a robust replacement. This thesis focuses on developing robust Bayesian models for uncertainty prediction to guide continuous and autonomous model development. First, neural network potentials built …
-
Building Reliable Machine Learning Interatomic Potentials for Catalyst Materials: Methods, Validation, and Applications
Machine learning interatomic potentials (MLIPs) are transforming atomistic simulation by approaching the accuracy of electronic structure methods while enabling molecular dynamics and structural optimization at much larger length and time scales. However, applying MLIPs to chemically complex …
-
Development of Interatomic Potentials with Uncertainty Quantification: Applications to Two-dimensional Materials
… science and engineering. Empirical interatomic potentials have proven to be an indis- pensable part of atomistic simulation due to their unrivaled computational efficiency in describing the interactions between atoms, which produce the forces governing atomic motion and deformation. Atomistic …
-
Parallel Tempering Monte Carlo Methods using Complex Potentials for Mono- and Binary Atomic Clusters
… the study of atomic clusters governed by complex potentials. We first examine second-shell neon clusters under extreme magnetic fields, in which magnetic-field-induced symmetry breaking leads to changes in ground-state structures. The 38-atom cluster transitions from an octahedral to a flatter …
-
Computational modeling of metallic nanoclusters and nano-alloys for catalytic and corrosion applications.
… novel Machine Learning methods, such as the Neural Network potentials. These are shown to clearly outperform standard empirical potentials. Afterwards, an analysis of the initial stages of the corrosion problem by means of ab-initio methods is performed. It is found that small amounts of Mg …
-
Desenvolvimento de potenciais interatômicos baseados em aprendizado de máquina aplicados a nanofilamentos de carbono
Nanofilamentos de carbono são nanomateriais unidimensionais constituídos por carbono e hidrogênio, que vêm ganhando muita notoriedade devido às suas excelentes propriedades mecânicas e potenciais aplicações tecnológicas. Trabalhos teóricocomputacionais, utilizando simulações de dinâmica molecular e …