Результаты исследований: Научные публикации в периодических изданиях › статья › Рецензирование
Реализация алгоритма оптимизации параметров молекулярно-динамического потенциала ReaxFF. / Шефов, К.С.; Степанова, М.М.
в: ПРОГРАММНЫЕ ПРОДУКТЫ И СИСТЕМЫ, № 2, 2014, стр. 25.Результаты исследований: Научные публикации в периодических изданиях › статья › Рецензирование
}
TY - JOUR
T1 - Реализация алгоритма оптимизации параметров молекулярно-динамического потенциала ReaxFF
AU - Шефов, К.С.
AU - Степанова, М.М.
PY - 2014
Y1 - 2014
N2 - Molecular dynamic methods that use ReaxFF (Reactive Force Field) allow obtaining sufficiently good results in simulating large multicomponent chemically reactive systems. ReaxFF is implemented in the standard software package LAMMPS. However, pre-optimization of numerous parameters of this force field for a specific system is required for correct computations. This paper presents a parallel program that implements one-parameter search algorithm for the molecular dy-namic force field ReaxFF. The program uses data obtained by accurate quantum chemical computations of characteristics (training set) of chemical compounds simple models for t he optimization process. The parameter search is based on the min-imization of the error function. This is a sum of squared deviations of training set entries computed by ReaxFF from ones computed by quantum chemical methods. The paper describes the parameter search algorithm and shows the program block-scheme. Molecular dynamic simulator LAMMPS is used to search the minimum o
AB - Molecular dynamic methods that use ReaxFF (Reactive Force Field) allow obtaining sufficiently good results in simulating large multicomponent chemically reactive systems. ReaxFF is implemented in the standard software package LAMMPS. However, pre-optimization of numerous parameters of this force field for a specific system is required for correct computations. This paper presents a parallel program that implements one-parameter search algorithm for the molecular dy-namic force field ReaxFF. The program uses data obtained by accurate quantum chemical computations of characteristics (training set) of chemical compounds simple models for t he optimization process. The parameter search is based on the min-imization of the error function. This is a sum of squared deviations of training set entries computed by ReaxFF from ones computed by quantum chemical methods. The paper describes the parameter search algorithm and shows the program block-scheme. Molecular dynamic simulator LAMMPS is used to search the minimum o
KW - numerical simulation
KW - chemically reactive systems
KW - reactive force field
KW - interaction potential
KW - parallel algorithm
KW - parameter optimization
KW - molecular dynamics
M3 - статья
SP - 25
JO - ПРОГРАММНЫЕ ПРОДУКТЫ И СИСТЕМЫ
JF - ПРОГРАММНЫЕ ПРОДУКТЫ И СИСТЕМЫ
SN - 0236-235X
IS - 2
ER -
ID: 5702251