Machine-Learning Accelerated Annealing with Fitting-Search Style for Multi-alloy Structure Predictions

TL;DR

This paper introduces CCOP, a machine-learning enhanced annealing method that significantly accelerates multi-alloy structure prediction by reducing computational costs and improving search efficiency using graph neural networks and reinforcement learning.

Contribution

The paper presents a novel crystal optimization program combining graph neural networks and reinforcement learning to improve the efficiency and accuracy of structural predictions in materials science.

Findings

CCOP reduces DFT computational cost by two orders of magnitude.

CCOP achieves high accuracy with minimal training data.

CCOP outperforms conventional methods in finding lowest-energy structures.

Abstract

Structural prediction for the discovery of novel materials is a long sought after goal of computational physics and materials sciences. The success is rather limited for methods such as the simulated annealing method (SA) that require expensive density functional theory (DFT) calculations and follow unintelligent search paths. Here a machine-learning based crystal combinatorial optimization program (CCOP) with a fitting-search style is proposed to drastically improve the efficiency of structural search in SA. CCOP uses a graph neural network energy prediction model to reduce the DFT cost and a deep reinforcement learning algorithm to direct the search path. Tests on six multi-alloys show the energy prediction model is capable of extracting the bonding characteristics of the complex alloys to achieve interpretability. It also achieves high accuracy with a tiny training set (an increment of 30 samples per iteration) by active learning in less than 5 iterations. Comparison with a few conventional methods shows that CCOP finds the lowest-energy structures with the smallest number of search steps. CCOP cuts the computing cost of SA by two orders of magnitude, while providing better search results than SA. CCOP is promising for serving as a broadly applicable tool for the efficient crystal structure predictions.