A sparsity preserving genetic algorithm for extracting diverse functional 3D designs from deep generative neural networks. (2020)

• Record Type: Journal Article
• Authors: Cunningham, James D.; Shu, Dule; Simpson, Timothy W.; Tucker, Conrad S.
• Is Part Of: Design science. Volume 6(2020)
• Publication Date: 2020
• Subjects: deep learning, -- computer-aided design, -- design space exploration, -- genetic algorithms; Design -- Research -- Periodicals; New products -- Management -- Periodicals; Design; Design -- Research; Electronic journals; Periodicals; 658.5752
• DOI: 10.1017/dsj.2020.9
• Languages: English
• ISSNs: 2053-4701

Abstract

Abstract : Generative neural networks (GNNs) have successfully used human-created designs to generate novel 3D models that combine concepts from disparate known solutions, which is an important aspect of design exploration. GNNs automatically learn a parameterization (or latent space ) of a design space, as opposed to alternative methods that manually define a parameterization. However, GNNs are typically not evaluated using an explicit notion of physical performance, which is a critical capability needed for design. This work bridges this gap by proposing a method to extract a set of functional designs from the latent space of a point cloud generating GNN, without sacrificing the aforementioned aspects of a GNN that are appealing for design exploration. We introduce a sparsity preserving cost function and initialization strategy for a genetic algorithm (GA) to optimize over the latent space of a point cloud generating autoencoder GNN. We examine two test cases, an example of generating ellipsoid point clouds subject to a simple performance criterion and a more complex example of extracting 3D designs with a low coefficient of drag. Our experiments show that the modified GA results in a diverse set of functionally superior designs while maintaining similarity to human-generated designs in the training data set.