GPR_calculator: An On-the-Fly Surrogate Model to Accelerate Massive Nudged Elastic Band Calculations

Abstract

We present GPR_calculator, a package based on Python and C++ programming languages to build an on-the-fly surrogate model using Gaussian Process Regression (GPR) to approximate expensive electronic structure calculations. The key idea is to dynamically train a GPR model during the simulation that can accurately predict energies and forces with uncertainty quantification. When the uncertainty is high, the expensive electronic structure calculation is performed to obtain the ground truth data, which is then used to update the GPR model. To illustrate the power of GPR_calculator, we demonstrate its application in Nudged Elastic Band (NEB) simulations of surface diffusion and reactions, achieving 3-10 times acceleration compared to pure ab initio calculations. The source code is available at https://github.com/MaterSim/GPR_calculator.

12 pages, 7 figures

References in corpus (6)

• Quantum ESPRESSO: a modular and open-source software project for quantum simulations of materials
• Array Programming with NumPy
• On-the-fly machine learning force field generation: Application to melting points
• Improved initial guess for minimum energy path calculations
• Nudged elastic band calculations accelerated with Gaussian process regression
• Spectral Neural Network Potentials for Binary Alloys

Cited by in corpus (1)

• Direct Deoxygenation of Phenol over Fe-based Bimetallic Surfaces using On-the-fly Surrogate Models