Neural reparameterization improves structural optimization

TL;DR

This paper introduces a neural network-based reparameterization technique for structural optimization, leading to improved design quality by leveraging implicit function bias, outperforming traditional methods on multiple tasks.

Contribution

It proposes using neural network parameterization for structural optimization, significantly enhancing solution quality over conventional density-based methods.

Findings

Achieved 50% more best designs than baseline methods.

Improved solution quality across 116 structural optimization tasks.

Demonstrated the effectiveness of neural reparameterization in structural design.

Abstract

Structural optimization is a popular method for designing objects such as bridge trusses, airplane wings, and optical devices. Unfortunately, the quality of solutions depends heavily on how the problem is parameterized. In this paper, we propose using the implicit bias over functions induced by neural networks to improve the parameterization of structural optimization. Rather than directly optimizing densities on a grid, we instead optimize the parameters of a neural network which outputs those densities. This reparameterization leads to different and often better solutions. On a selection of 116 structural optimization tasks, our approach produces the best design 50% more often than the best baseline method.