Diffusion model-based inverse design for thermal transparency

Bin Liu,Liujun Xu,Yixi Wang,Jiping Huang

DOI: https://doi.org/10.1063/5.0197999

IF: 2.877

2024-03-25

Journal of Applied Physics

Abstract:Generative models in the field of artificial intelligence and their applications and deployment have demonstrated their great strength in the past few years. Of the vast spectrum of generative models, diffusion probabilistic models have proven to be particularly powerful and productive, transforming notions such as text-to-image and text-to-video generation from ideas into practical applications. In our previous works, we proposed a thermal metamaterial-based periodic interparticle interaction mechanism for heat management, with a specific application in thermal transparency. To address the challenging problems associated with the inverse design of thermal metamaterial structures, we employed an autoencoder-based machine learning approach and a reinforcement learning-based approach successfully. In this work, we demonstrate that our particular problems with the inverse design of thermal metamaterial-based periodic lattices for the realization of thermal transparency can also be reframed and efficiently solved by training a generative diffusion probabilistic model that can generate the design parameters corresponding to the desired response. Furthermore, we show that for a specific response, multiple sets of design parameters can be obtained by simply performing multiple inferences with the generative diffusion probabilistic model, enabling us to select the ones that can be more economical to fabricate and implement. Our work is among the first to use a diffusion model for the inverse design of thermal metamaterial-based structures and demonstrates the effectiveness of generating low-dimensional design parameters through a diffusion model.

physics, applied

What problem does this paper attempt to address?

### Problems the Paper Aims to Solve

This paper is dedicated to solving the inverse design problem of thermal transparency materials. Specifically, the authors utilize diffusion models to achieve the inverse design of periodic lattice structures based on thermal metamaterials.

#### Main Objectives:
1. **Inverse Design Problem**: By training a generative diffusion probabilistic model to obtain design parameters corresponding to the desired thermal transfer behavior. Traditional methods have limitations in handling thermal transparency materials, especially due to application barriers caused by asymmetric interactions. This paper proposes a new approach, using diffusion models to redefine and efficiently solve this problem.
2. **Multiple Solutions Problem**: By performing multiple inferences, various sets of design parameters can be obtained, allowing for the selection of more economical or feasible design schemes in practical applications. This method addresses the one-to-many mapping problem from design space to response space and transforms it into an advantageous feature.

#### Specific Approaches:
- **Data Preparation**: Generating a dataset containing a large number of random design parameters through finite element method (FEM) simulations.
- **Model Training**: Using an autoencoder to compress high-dimensional response vectors, then training a diffusion model to generate the corresponding design parameters.
- **Noise Prediction Network**: Designing a noise prediction neural network architecture to optimize model performance and avoid overfitting.

Through these methods, the paper demonstrates the effectiveness and potential of diffusion models in solving the inverse design problem of thermal metamaterials.