Flow Battery Manifold Design with Heterogeneous Inputs Through Generative Adversarial Neural Networks

2025/08/12 by Eric Seng, Seng, Eric, Hugh O’Connor +7
Computer Science · Materials Science · #Advanced Multi-Objective Optimization Algorithms #Advanced Neural Network Applications #Applications (stat.AP) #FOS: Computer and information sciences #Machine Learning (cs.LG) #Machine Learning in Materials Science

paper · pdf · doi:10.48550/arxiv.2508.08863

openalex publication_date 2025/08/12 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

Generative machine learning has emerged as a powerful tool for design representation and exploration. However, its application is often constrained by the need for large datasets of existing designs and the lack of interpretability about what features drive optimality. To address these challenges, we introduce a systematic framework for constructing training datasets tailored to generative models and demonstrate how these models can be leveraged for interpretable design. The novelty of this work is twofold: (i) we present a systematic framework for generating archetypes with internally homogeneous but mutually heterogeneous inputs that can be used to generate a training dataset, and (ii) we show how integrating generative models with Bayesian optimization can enhance the interpretability of the latent space of admissible designs. These findings are validated by using the framework to design a flow battery manifold, demonstrating that it effectively captures the space of feasible designs, including novel configurations while enabling efficient exploration. This work broadens the applicability of generative machine-learning models in system designs by enhancing quality and reliability.

Citations

• Operator Learning with Gaussian Processes
• Fairness- and uncertainty-aware data generation for data-driven design
• Diffusing the Optimal Topology: A Generative Optimization Approach
• Hidden-Variables Genetic Algorithm for Variable-Size Design Space Optimal Layout Problems with Application to Aerospace Vehicles
• Hierarchical Text-Conditional Image Generation with CLIP Latents
• Deep Generative Models in Engineering Design: A Review
• Generative Adversarial Networks
• Data-Driven Multiscale Design of Cellular Composites with Multiclass Microstructures for Natural Frequency Maximization
• BIKED: A Dataset for Computational Bicycle Design with Machine Learning Benchmarks
• Airfoil Design Parameterization and Optimization using Bézier Generative Adversarial Networks
• Denoising Diffusion Probabilistic Models
• Training Generative Adversarial Networks with Limited Data
• Autoencoding Generative Adversarial Networks
• Bayesian optimization of variable-size design space problems
• A Latent Variable Approach to Gaussian Process Modeling with Qualitative and Quantitative Factors
• InfoGAN: Interpretable Representation Learning by Information Maximizing Generative Adversarial Nets
• Conditional Generative Adversarial Nets
• Entropy Search for Information-Efficient Global Optimization

Related

• PaDGAN: A Generative Adversarial Network for Performance Augmented Diverse Designs
• Generative VS non-Generative Models in Engineering Shape Optimization
• Beyond Statistical Similarity: Rethinking Metrics for Deep Generative Models in Engineering Design
• Towards Goal, Feasibility, and Diversity-Oriented Deep Generative Models in Design
• Deep Generative Models in Engineering Design: A Review