Post-graduate theses

Current Record: 4602 of 6695

[Add to Basket]
Identifier	000412825
Title	Stochastic and deterministic parametrization of coarse graining models for molecular systems
Alternative Title	Στοχαστική και ντετερμινιστική παραμετροποίηση μοντέλων αδροποίησης για μοριακά συστήματα
Author	Χαζηράκης, Αντώνιος
Thesis advisor	Χαρμανδάρης, Βαγγέλης
Reviewer	Ζουράρης, Γεώργιος Καλλιγιανάκη, Ευαγγελία
Abstract	Molecular dynamics (MD) simulations study the physical movement of atoms, providing valuable insight for a variety of physical systems. Despite of the modern advances to the available computational resources, the length and time scales of the physical systems un- der study are still limited. The only remedy is the use of mathematical methods in order to reduce the dimensionality of the physical system under study. The most often used methodology for doing that is Coarse Graining (CG). Coarse graining methodologies have both mathematical as well as computational and numerical challenges. We experi- ment with three different methods of coarse graining, the Iterative Boltzmann Inversion (IBI) method, the Force Matching (FM or MSCG) method and the Relative Entropy (RE) method, for a variety of physical systems and compare the results. The IBI method finds the CG potential that reproduces the atomistic pair correlation function. The FM method finds the CG potential that minimizes the mean distance between the atomistic and the CG forces. The RE method finds the potential that minimizes the difference between the atomistic and CG coordinate probability distributions. The numerical treat- ment of the IBI and FM methods is deterministic, although it involves calculation of expectations. The RE method not only involves calculation of expectations, but also performs minimization using the stochastic Robins-Monro scheme.
Language	English
Issue date	2017-11-24
Collection	School/Department--School of Sciences and Engineering--Department of Mathematics and Applied Mathematics--Post-graduate theses
	Type of Work--Post-graduate theses
Permanent Link	https://elocus.lib.uoc.gr//dlib/d/a/8/metadata-dlib-1512034410-566669-5277.tkl
Views	717

Digital Documents
	Download document View document Views : 28