Abstract:Generative models have had a profound impact on vision and language, paving the way for a new era of multimodal generative applications. While these successes have inspired researchers to explore using generative models in science and engineering to accelerate the design process and reduce the reliance on iterative optimization, challenges remain. Specifically, engineering optimization methods based on physics still outperform generative models when dealing with constrained environments where data is scarce and precision is paramount. To address these challenges, we introduce Diffusion Optimization Models (DOM) and Trajectory Alignment (TA), a learning framework that demonstrates the efficacy of aligning the sampling trajectory of diffusion models with the optimization trajectory derived from traditional physics-based methods. This alignment ensures that the sampling process remains grounded in the underlying physical principles. Our method allows for generating feasible and high-performance designs in as few as two steps without the need for expensive preprocessing, external surrogate models, or additional labeled data. We apply our framework to structural topology optimization, a fundamental problem in mechanical design, evaluating its performance on in- and out-of-distribution configurations. Our results demonstrate that TA outperforms state-of-the-art deep generative models on in-distribution configurations and halves the inference computational cost. When coupled with a few steps of optimization, it also improves manufacturability for out-of-distribution conditions. By significantly improving performance and inference efficiency, DOM enables us to generate high-quality designs in just a few steps and guide them toward regions of high performance and manufacturability, paving the way for the widespread application of generative models in large-scale data-driven design.

What problem does this paper attempt to address?

The paper attempts to address the problem of how to utilize generative models to accelerate the design process in engineering design, particularly in the field of Structural Topology Optimization, while maintaining or improving the quality and manufacturability of the designs. Despite the significant progress made by existing generative models in visual and language processing, and the inspiration these successes have provided for researchers to explore the application of generative models in science and engineering, they still face challenges when dealing with problems in constrained environments. Specifically, physics-based engineering optimization methods still outperform generative models in situations where data is scarce and high precision is required. The paper proposes a new framework—Diffusion Optimization Models (DOM) and Trajectory Alignment (TA), aimed at ensuring that the sampling process is rooted in underlying physical principles by aligning the sampling trajectories of diffusion models with the optimization trajectories obtained from traditional physics-based methods. This approach allows for the generation of feasible and high-performance designs within very few steps (e.g., 2 steps) without the need for expensive preprocessing, external surrogate models, or additional labeled data. By applying the framework to structural topology optimization problems, the paper evaluates its performance, demonstrating the superiority of trajectory alignment in in-distribution configurations and its effectiveness in reducing inference computational costs. Furthermore, when combined with a few optimization steps, this method can also improve manufacturability under out-of-distribution conditions. In summary, the main goal of the paper is to develop an efficient and effective method that combines data-driven and physics-based approaches to solve complex engineering design problems, particularly those that need to meet stringent constraints.

Aligning Optimization Trajectories with Diffusion Models for Constrained Design Generation

Enhancing Topology Optimization with Adaptive Deep Learning

Texture-guided Generative Structural Designs under Local Control

Diffusing the Optimal Topology: A Generative Optimization Approach

Guided Trajectory Generation with Diffusion Models for Offline Model-based Optimization

Model-Based Diffusion for Trajectory Optimization

Diffusion Models Beat GANs on Topology Optimization

Optimizing Diffusion Models for Joint Trajectory Prediction and Controllable Generation

Constraint-Aware Diffusion Models for Trajectory Optimization

Multidisciplinary Topology Optimization Using Generative Adversarial Networks for Physics-Based Design Enhancement

Geometric Trajectory Diffusion Models

Bridging Model-Based Optimization and Generative Modeling via Conservative Fine-Tuning of Diffusion Models

Equality Constrained Diffusion for Direct Trajectory Optimization

Towards Goal, Feasibility, and Diversity-Oriented Deep Generative Models in Design

Latent Diffusion Models for Structural Component Design

Stylish and Functional: Guided Interpolation Subject to Physical Constraints

Constrained Synthesis with Projected Diffusion Models

Generative Transfer Optimization for Aerodynamic Design

Shape related constraints aware generation of Mechanical Designs through Deep Convolutional GAN

Scalable Deep-Learning-Accelerated Topology Optimization for Additively Manufactured Materials

Dynamics-Guided Diffusion Model for Robot Manipulator Design