A fast and flexible algorithm for microstructure reconstruction combining simulated annealing and deep learning

TL;DR

This paper introduces a novel microstructure reconstruction algorithm combining simulated annealing and deep learning, achieving faster, flexible, and efficient reconstructions from minimal data, suitable for various material types.

Contribution

The paper presents a new neural network method based on an improved simulated annealing framework that is faster and more adaptable than existing algorithms.

Findings

Reconstruction can be completed in a short time with only one 2D image.

The algorithm can reconstruct arbitrary size microstructures.

Effective on both isotropic and anisotropic materials.

Abstract

The microstructure analyses of porous media have considerable research value for the study of macroscopic properties. As the premise of conducting these analyses, the accurate reconstruction of microstructure digital model is also an important component of the research. Computational reconstruction algorithms of microstructure have attracted much attention due to their low cost and excellent performance. However, it is still a challenge for computational reconstruction algorithms to achieve faster and more efficient reconstruction. The bottleneck lies in computational reconstruction algorithms, they are either too slow (traditional reconstruction algorithms) or not flexible to the training process (deep learning reconstruction algorithms). To address these limitations, we proposed a fast and flexible computational reconstruction algorithm, neural networks based on improved simulated annealing framework (ISAF-NN). The proposed algorithm is flexible and can complete training and reconstruction in a short time with only one two-dimensional image. By adjusting the size of input, it can also achieve reconstruction of arbitrary size. Finally, the proposed algorithm is experimentally performed on a variety of isotropic and anisotropic materials to verify the effectiveness and generalization.