Denoising diffusion models for inverse design of inflatable structures with programmable deformations

2025-08-18

AI Summary

This study addresses the inverse design of inflatable structures under pneumatic actuation, targeting large-deformation configurations. We propose a generative framework based on Denoising Diffusion Probabilistic Models (DDPMs), formulated as a conditional generation task: given a geometric descriptor—scalar or high-dimensional—of the target deformed shape, the model directly synthesizes both the initial undeformed configuration and a spatially varying material distribution satisfying functional requirements. Our approach innovatively integrates image-based representation, finite-element simulation feedback, and geometric descriptor extraction to enable concurrent multi-solution and diverse design under complex physical constraints. Experimental results demonstrate that the method generates high-fidelity candidate designs within seconds, achieving low post-inflation shape error and high morphological fidelity to the target deformation. This significantly enhances both design efficiency and controllability for programmable soft structures.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Search and Optimization: Sampling/Simulation-based Search
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language

Application Category

• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs

Abstract

Programmable structures are systems whose undeformed geometries and material property distributions are deliberately designed to achieve prescribed deformed configurations under specific loading conditions. Inflatable structures are a prominent example, using internal pressurization to realize large, nonlinear deformations in applications ranging from soft robotics and deployable aerospace systems to biomedical devices and adaptive architecture. We present a generative design framework based on denoising diffusion probabilistic models (DDPMs) for the inverse design of elastic structures undergoing large, nonlinear deformations under pressure-driven actuation. The method formulates the inverse design as a conditional generation task, using geometric descriptors of target deformed states as inputs and outputting image-based representations of the undeformed configuration. Representing these configurations as simple images is achieved by establishing a pre- and postprocessing pipeline that involves a fixed image processing, simulation setup, and descriptor extraction methods. Numerical experiments with scalar and higher-dimensional descriptors show that the framework can quickly produce diverse undeformed configurations that achieve the desired deformations when inflated, enabling parallel exploration of viable design candidates while accommodating complex constraints.

Problem

• Inverse design of inflatable structures with programmable deformations
• Generative framework for pressure-driven nonlinear deformation structures
• Conditional generation of undeformed configurations from target deformations

Innovation

• Denoising diffusion models for inverse design
• Conditional generation using geometric descriptors
• Image-based preprocessing and simulation pipeline