Knowledge-Constrained Shape Optimization with a Mixture-of-Experts Neural Operator for High-Confidence Design
Quick Answer
The proposed knowledge-constrained shape optimization framework utilizes a Mixture-of-Experts Neural Operator (MoE-NO) to enhance drag prediction accuracy to 94.34% while achieving a test-set MAPE of 1.16%.
Quick Take
The proposed knowledge-constrained shape optimization framework utilizes a Mixture-of-Experts Neural Operator (MoE-NO) to enhance drag prediction accuracy to 94.34% while achieving a test-set MAPE of 1.16%. This approach enables reliable optimization in aerodynamic design, resulting in drag coefficient reductions of 4% to 10% across various vehicle models.
Key Points
- MoE-NO improves drag prediction and trend consistency in heterogeneous aerodynamic datasets.
- Achieved a test-set MAPE of 1.16% and trend-prediction accuracy of 94.34%.
- Drag coefficient reductions of approximately 4% to 10% validated by CFD experiments.
- Framework translates user intent into quantifiable parameters for controlled optimization.
- Uncertainty estimation detects out-of-distribution geometries for enhanced sample enrichment.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Engineering shape optimization faces challenges in both expert-dependent problem setup and surrogate-model reliability. In practical aerodynamic design, optimization settings such as editable regions, deformation ranges, and design-preservation constraints are typically specified manually by experienced engineers, while surrogate-based optimization may become unreliable for heterogeneous geometry databases and out-of-distribution designs. To address these challenges, we propose a knowledge-constrained shape-optimization framework that translates knowledge-based constraints and user intent into quantifiable parameters of DFFD-based deformation operators, enabling engineering-aware and controllable constrained optimization. We further develop a Mixture-of-Experts Neural Operator (MoE-NO) to improve drag prediction and trend consistency over heterogeneous aerodynamic datasets. Based on the MoE-NO encoder and Mahalanobis distance, an uncertainty-estimation strategy is introduced to detect out-of-distribution geometries and selectively trigger physics-solver feedback for local sample enrichment. Experiments on in-house MPV, SUV, and Sedan datasets show that MoE-NO achieves a test-set MAPE of $1.16\%$ and a trend-prediction accuracy of $94.34\%$, outperforming the best baseline results of $1.52\%$ and $90.34\%$, respectively. Vehicle shape-optimization experiments further yield CFD-validated drag coefficient reductions of approximately $4\%$ to $10\%$.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Computational Physics (physics.comp-ph) |
| Cite as: | arXiv:2607.09763 [cs.CV] |
| (or arXiv:2607.09763v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2607.09763 arXiv-issued DOI via DataCite |
Submission history
From: Yuanwei Bin [view email]
[v1]
Tue, 7 Jul 2026 01:29:09 UTC (8,938 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.