Design Strategy Network: A deep hierarchical framework to represent generative design strategies in complex action spaces

TL;DR

This paper introduces the Design Strategy Network, a hierarchical deep learning framework that effectively models complex, diverse action spaces in generative design tasks, outperforming non-hierarchical methods.

Contribution

The paper presents a novel hierarchical deep learning framework that captures complex design strategies in diverse action spaces, including hybrid and state-dependent constraints.

Findings

DSN outperforms non-hierarchical methods in complex action spaces

Successfully predicts human designer actions in truss design

Hierarchical decomposition improves policy learning in generative design

Abstract

Generative design problems often encompass complex action spaces that may be divergent over time, contain state-dependent constraints, or involve hybrid (discrete and continuous) domains. To address those challenges, this work introduces Design Strategy Network (DSN), a data-driven deep hierarchical framework that can learn strategies over these arbitrary complex action spaces. The hierarchical architecture decomposes every action decision into first predicting a preferred spatial region in the design space and then outputting a probability distribution over a set of possible actions from that region. This framework comprises a convolutional encoder to work with image-based design state representations, a multi-layer perceptron to predict a spatial region, and a weight-sharing network to generate a probability distribution over unordered set-based inputs of feasible actions. Applied to a truss design study, the framework learns to predict the actions of human designers in the study, capturing their truss generation strategies in the process. Results show that DSNs significantly outperform non-hierarchical methods of policy representation, demonstrating their superiority in complex action space problems.