WindMiL: Equivariant Graph Learning for Wind Loading Prediction

2025-11-02

AI Summary

Accurate prediction of wind loads on buildings is critical for structural safety and sustainable design, yet conventional methods—wind tunnel testing and large-eddy simulation (LES)—are computationally prohibitive (≥24 hours per case), hindering large-scale parametric studies. To address this, we propose a reflection-equivariant graph neural network (GNN) surrogate model that integrates roof geometry representations derived from signed distance function interpolation with an LES dataset, explicitly encoding mirror-symmetry constraints to ensure physical consistency. Our method achieves superior generalization and robustness: root-mean-square error (RMSE) in mean pressure coefficients ≤ 0.02, and mirror-symmetry test accuracy exceeding 96%—substantially outperforming non-equivariant baselines. This work establishes a new paradigm for efficient, interpretable, and scalable wind load prediction.

Technology Category

• Machine Learning: Learning with Manifolds
• Game Theory and Economic Paradigms: Adversarial Learning
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Semantics and Knowledge: Methods, algorithms and applications for the development of semantic models, knowledge graphs and other forms of structured data models with machine-interpretable semantics

Abstract

Accurate prediction of wind loading on buildings is crucial for structural safety and sustainable design, yet conventional approaches such as wind tunnel testing and large-eddy simulation (LES) are prohibitively expensive for large-scale exploration. Each LES case typically requires at least 24 hours of computation, making comprehensive parametric studies infeasible. We introduce WindMiL, a new machine learning framework that combines systematic dataset generation with symmetry-aware graph neural networks (GNNs). First, we introduce a large-scale dataset of wind loads on low-rise buildings by applying signed distance function interpolation to roof geometries and simulating 462 cases with LES across varying shapes and wind directions. Second, we develop a reflection-equivariant GNN that guarantees physically consistent predictions under mirrored geometries. Across interpolation and extrapolation evaluations, WindMiL achieves high accuracy for both the mean and the standard deviation of surface pressure coefficients (e.g., RMSE $leq 0.02$ for mean $C_p$) and remains accurate under reflected-test evaluation, maintaining hit rates above $96%$ where the non-equivariant baseline model drops by more than $10%$. By pairing a systematic dataset with an equivariant surrogate, WindMiL enables efficient, scalable, and accurate predictions of wind loads on buildings.

Problem

• Predicting wind loading on buildings efficiently and accurately
• Overcoming computational expense of traditional simulation methods
• Developing symmetry-aware machine learning for physical consistency

Innovation

• Equivariant graph neural networks for wind prediction
• Large-scale dataset generation via LES simulations
• Reflection-equivariant GNN ensuring physical consistency