TopologyGAN: Topology Optimization Using Generative Adversarial Networks Based on Physical Fields Over the Initial Domain
This addresses the challenge of encoding design problems in topology optimization for engineers, though it is incremental as it builds on existing network models.
The paper tackles the problem of poor generalization in deep learning-based topology optimization by proposing TopologyGAN, which uses physical fields over the initial domain as inputs to a cGAN, resulting in a nearly 3x reduction in mean squared error and a 2.5x reduction in mean absolute error on unseen boundary conditions.
In topology optimization using deep learning, load and boundary conditions represented as vectors or sparse matrices often miss the opportunity to encode a rich view of the design problem, leading to less than ideal generalization results. We propose a new data-driven topology optimization model called TopologyGAN that takes advantage of various physical fields computed on the original, unoptimized material domain, as inputs to the generator of a conditional generative adversarial network (cGAN). Compared to a baseline cGAN, TopologyGAN achieves a nearly $3\times$ reduction in the mean squared error and a $2.5\times$ reduction in the mean absolute error on test problems involving previously unseen boundary conditions. Built on several existing network models, we also introduce a hybrid network called U-SE(Squeeze-and-Excitation)-ResNet for the generator that further increases the overall accuracy. We publicly share our full implementation and trained network.