Dynamic Learning of Correlation Potentials for a Time-Dependent Kohn-Sham System

We develop methods to learn the correlation potential for a time-dependent Kohn-Sham (TDKS) system in one spatial dimension. We start from a low-dimensional two-electron system for which we can numerically solve the time-dependent Schrödinger equation; this yields electron densities suitable for training models of the correlation potential. We frame the learning problem as one of optimizing a least-squares objective subject to the constraint that the dynamics obey the TDKS equation. Applying adjoints, we develop efficient methods to compute gradients and thereby learn models of the correlation potential. Our results show that it is possible to learn values of the correlation potential such that the resulting electron densities match ground truth densities. We also show how to learn correlation potential functionals with memory, demonstrating one such model that yields reasonable results for trajectories outside the training set.

[1]  E. K. U. Gross,et al.  Density-Functional Theory for Time-Dependent Systems , 1984 .

[2]  N. Maitra,et al.  Perspective: Fundamental aspects of time-dependent density functional theory. , 2016, The Journal of chemical physics.

[3]  Á. Rubio,et al.  Universal dynamical steps in the exact time-dependent exchange-correlation potential. , 2012, Physical review letters.

[4]  Angel Rubio,et al.  Propagators for the time-dependent Kohn-Sham equations. , 2004, The Journal of chemical physics.

[5]  Ryan J. McCarty,et al.  Learning to Approximate Density Functionals. , 2021, Accounts of chemical research.

[6]  A. Borzì.  Quantum optimal control using the adjoint method , 2012 .

[7]  Alfio Borzì,et al.  A Theoretical Investigation of Time-Dependent Kohn-Sham Equations , 2017, SIAM J. Math. Anal..

[8]  Sepp Hochreiter,et al.  Self-Normalizing Neural Networks , 2017, NIPS.

[9]  George M. Siouris,et al.  Applied Optimal Control: Optimization, Estimation, and Control , 1979, IEEE Transactions on Systems, Man, and Cybernetics.

[10]  Carsten A. Ullrich,et al.  Time-Dependent Density-Functional Theory: Concepts and Applications , 2012 .

[11]  E. Gross,et al.  Optimal control theory for quantum-classical systems: Ehrenfest molecular dynamics based on time-dependent density-functional theory , 2013, 1308.4162.

[12]  R. Nagai,et al.  Machine learning exchange-correlation potential in time-dependent density-functional theory , 2020, 2002.06542.

[13]  L. Hasdorff.  Gradient Optimization and Nonlinear Control , 1976 .

[14]  E K U Gross,et al.  Controlling the dynamics of many-electron systems from first principles: a combination of optimal control and time-dependent density-functional theory. , 2010, Physical review letters.

[15]  M. Feit,et al.  Solution of the Schrödinger equation by a spectral method , 1982 .

[16]  J. R. Morris,et al.  Time-dependent propagation of high energy laser beams through the atmosphere , 1976 .

[17]  Kieron Burke,et al.  Memory in time-dependent density functional theory. , 2002, Physical review letters.

[18]  Ghosh,et al.  Density-functional theory for time-dependent systems. , 1987, Physical review. A, General physics.

[19]  Ryo Nagai,et al.  Neural-network Kohn-Sham exchange-correlation potential and its out-of-training transferability. , 2018, The Journal of chemical physics.

[20]  R. Leeuwen,et al.  MAPPING FROM DENSITIES TO POTENTIALS IN TIME-DEPENDENT DENSITY-FUNCTIONAL THEORY , 1999 .

[21]  A. Borzì,et al.  Optimal control with TDKS , 2018 .