arXiv · 1610.08423
Co-iterative augmented Hessian method for orbital optimization
Abstract
Orbital optimization procedure is widely called in electronic structure simulation. To efficiently find the orbital optimization solution, we developed a new second order orbital optimization algorithm, co-iteration augmented Hessian (CIAH) method. In this method, the orbital optimization is embedded in the diagonalization procedure for augmented Hessian (AH) eigenvalue equation. Approximations for Hessian matrix can be easily employed in this method to improve the computational costs. We numerically performed the CIAH algorithm with SCF convergence of 20 challenging systems and Boys localization of C60 molecule. We found that CIAH algorithm has better SCF convergence and less computational costs than direct inversion iterative subspace (DIIS) algorithm. The numerical tests suggest that CIAH is a stable, reliable and efficient algorithm for orbital optimization problem.
Explore related subjects
Keep this discovery
Qiming Sun. 2016-10-26. Co-iterative augmented Hessian method for orbital optimization. https://arxiv.org/abs/1610.08423
Cite the original work for its findings. Save a collection to share your selection of sources.