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

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.

Exploiting Linear Interpolation of Variational Autoencoders for Satisfying Preferences in Evolutionary Design Optimization

“…Thus, in addition, Saha et al [8] utilized a point cloud-based variational autoencoder (PC-VAE) [8] that builds on the point cloud autoencoder originally proposed in [1] and evaluated the model for novel and realistic 3D shape generation for engineering applications. In the context of optimization tasks, previous research [9]- [11] utilized the low-dimensional latent representation of an AE or a VAE for single and multiobjective optimization. Due to better generative abilities of PC-VAE in comparison to AE [1], we intend to use the latent representation of the PC-VAE proposed in [8] as decision variables for our optimization task.…”

Design representation for performance evaluation of 3D shapes in structure-aware generative design

“…In 3D design applications, Shu et al (2020) present a method that combines GANs and a physics-based virtual environment introduced by Dering et al (2018) to generate high-performance 3D aircraft models. Zhang et al (2019) propose a method using VAEs, a physics-based simulator, and a functional design optimizer to synthesize 3D aircraft with prescribed engineering performance.…”

“…Such a latent vector space is a low-dimensional representation of the design space from which the data were observed. Since the training process combines features from all existing designs, new designs that are not seen from existing data can be sampled from the latent design space (Krish 2011; Cunningham et al 2020). Therefore, DDGD methods have become an important tool for the generation of conceptual design ideas due to their ability to quickly generate a large number of novel designs.…”

“…Traditionally, most DDGD methods treat each design data as a monolithic whole (i.e. one single object without considering the interconnections of components as shown in Figure 1(a)) for model training (Shu et al 2020). In this paper, we refer to them as traditional DDGD methods.…”