ML LCOUPLE: Difference between revisions

From VASP Wiki
m (Karsai moved page ML FF LCOUPLE MB to ML LCOUPLE)
No edit summary
Line 1: Line 1:
{{TAGDEF|ML_FF_LCOUPLE_MB|[logical]|.FALSE.}}
{{TAGDEF|ML_LCOUPLE|[logical]|.FALSE.}}


Description: This tag specifies whether coupling parameters are used for the calculation of chemical potentials or not within the machine learning force field method.
Description: This tag specifies whether coupling parameters are used for the calculation of chemical potentials or not within the machine learning force field method.
Line 26: Line 26:


For thermodynamic integration the following parameters have to be set:
For thermodynamic integration the following parameters have to be set:
*{{TAG|ML_FF_ISTART}}=2.
*{{TAG|ML_ISTART}}=2.
*{{TAG|ML_FF_LCOUPLE_MB}}=''.TRUE.''.
*{{TAG|ML_LCOUPLE}}=''.TRUE.''.
*The number of atoms for which a coupling parameter is introduced (<math>i \notin M </math>): {{TAG|ML_FF_NATOM_COUPLED_MB}}.
*The number of atoms for which a coupling parameter is introduced (<math>i \notin M </math>): {{TAG|ML_NATOM_COUPLED}}.
*The list of atom indices that for that the coupling parameter is applied in the interaction: {{TAG|ML_FF_ICOUPLE_MB}}.
*The list of atom indices that for that the coupling parameter is applied in the interaction: {{TAG|ML_ICOUPLE}}.
*The strength of the coupling parameter <math>lambda</math> between 0 and 1: {{TAG|ML_FF_RCOUPLE_MB}}.
*The strength of the coupling parameter <math>lambda</math> between 0 and 1: {{TAG|ML_RCOUPLE}}.


The derivative of the hamiltonian with respect to the coupling constant <math>dH/d\lambda</math> is written out at every MD step to the {{TAG|ML_LOGFILE}}. A sample output should look like the following
The derivative of the hamiltonian with respect to the coupling constant <math>dH/d\lambda</math> is written out at every MD step to the {{TAG|ML_LOGFILE}}. A sample output should look like the following
Line 41: Line 41:


== Related Tags and Sections ==
== Related Tags and Sections ==
{{TAG|ML_FF_LMLFF}}, {{TAG|ML_FF_NATOM_COUPLED_MB}}, {{TAG|ML_FF_ICOUPLE_MB}}, {{TAG|ML_FF_RCOUPLE_MB}}
{{TAG|ML_LMLFF}}, {{TAG|ML_NATOM_COUPLED}}, {{TAG|ML_ICOUPLE}}, {{TAG|ML_RCOUPLE}}


{{sc|ML_FF_LCOUPLE_MB|Examples|Examples that use this tag}}
{{sc|ML_LCOUPLE|Examples|Examples that use this tag}}
----
----


[[Category:INCAR]][[Category:Machine Learning]][[Category:Machine Learned Force Fields]][[Category: Alpha]]
[[Category:INCAR]][[Category:Machine Learning]][[Category:Machine Learned Force Fields]][[Category: Alpha]]

Revision as of 08:53, 23 August 2021

ML_LCOUPLE = [logical]
Default: ML_LCOUPLE = .FALSE. 

Description: This tag specifies whether coupling parameters are used for the calculation of chemical potentials or not within the machine learning force field method.


In thermodynamic integration a coupling parameter [math]\displaystyle{ \lambda }[/math] is introduced to the Hamiltonian to smoothly switch between a "non-interacting" reference state and a "fully-interacting" state. The change of the free energy along this path is written as

[math]\displaystyle{ \Delta \mu = \int\limits_{0}^{1} \langle \frac{dH(\lambda)}{d\lambda} \rangle_{\lambda} d\lambda. }[/math]

Using machine learning force fields the Hamiltonian can be written as

[math]\displaystyle{ H (\lambda) = \sum\limits_{i=1}^{N_{a}} \frac{|\mathbf{p}_{i}|^2}{2m_{i}} + \sum\limits_{i \notin M} U_{i}(\lambda) + \lambda \sum\limits_{i \in M} U_{i}(\lambda) + \sum\limits_{i}^{N_{a}} U_{i,\mathbf{atom}}. }[/math]

where [math]\displaystyle{ N_{a} }[/math] denotes the number of atoms and [math]\displaystyle{ U_{i,\mathbf{atom}} }[/math] is an atomic reference energy for a single non interacting atom. The first term in the equation describes the potential energy and the second and third term describe the potential energy of an atom [math]\displaystyle{ i }[/math]. The index [math]\displaystyle{ M }[/math] denotes the atoms whose interaction is controlled by a coupling parameter. The interaction of the atoms are controlled by scaling the contributions to the atom density via the coupling parameter

[math]\displaystyle{ \rho (\mathbf{r},\lambda) = \sum\limits_{j\notin M} f_{\mathrm{cut}} \left( \left| \mathbf{r}_{j} - \mathbf{r}_{i} \right| \right) g \left[ \mathbf{r} - \left( \mathbf{r}_{j} - \mathbf{r}_{i} \right) \right] + \lambda \sum\limits_{j\in M} f_{\mathrm{cut}} \left( \left| \mathbf{r}_{j} - \mathbf{r}_{i} \right| \right) g \left[ \mathbf{r} - \left( \mathbf{r}_{j} - \mathbf{r}_{i} \right) \right]. }[/math]


Further details on the implementation can be found in reference [1].

For thermodynamic integration the following parameters have to be set:

  • ML_ISTART=2.
  • ML_LCOUPLE=.TRUE..
  • The number of atoms for which a coupling parameter is introduced ([math]\displaystyle{ i \notin M }[/math]): ML_NATOM_COUPLED.
  • The list of atom indices that for that the coupling parameter is applied in the interaction: ML_ICOUPLE.
  • The strength of the coupling parameter [math]\displaystyle{ lambda }[/math] between 0 and 1: ML_RCOUPLE.

The derivative of the hamiltonian with respect to the coupling constant [math]\displaystyle{ dH/d\lambda }[/math] is written out at every MD step to the ML_LOGFILE. A sample output should look like the following

dH/dRCOUPLE (eV):     0.893558

References


Related Tags and Sections

ML_LMLFF, ML_NATOM_COUPLED, ML_ICOUPLE, ML_RCOUPLE

Examples that use this tag