发表机构
State Key Laboratory of Synergistic Chem-Bio Synthesis; Department of Chemical Engineering, School of Chemistry and Chemical Engineering; Shanghai Jiao Tong University（协同化学生物合成国家重点实验室; 化学工程系，化学与化工学院; 上海交通大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本文针对 NACA4418 翼型流动，通过输入干预和压力测试证明神经算子预测误差与物理统计误差不一致，提出应补充物理诊断指标。
AI 中文摘要
机器学习替代模型加速了物理模拟，但更低的预测误差未必对应于物理相关流动统计量中更低的误差。我们使用成对的计算流体动力学模拟和实验粒子图像测速测量，针对 NACA4418 翼型周围的流动研究了这一问题。一种保持均值的输入干预从观测流动历史的选定区域中去除速度波动。在四种神经算子中，从最活跃的 10% 有效观测单元中去除波动比等面积随机去除对预测的改变更大。由于掩码未针对去除的波动能量进行匹配，这种对比衡量的是敏感性，而非物理重要性的独立证据。另外，一个 CNO 在速度场误差上更低，但在两个分析子集上的双分量波动能量误差显著高于参考。输出衰减压力测试也证明了基准误差与域求和波动能量之间的不一致。这些单一基准结果促使在聚合预测分数之外报告补充的物理诊断；它们并未确立反事实的物理正确性。
英文摘要
Machine-learning surrogates accelerate physical simulation, but lower prediction error need not coincide with lower error in physically relevant flow statistics. We examine this question for flow around a NACA4418 airfoil using paired computational-fluid-dynamics simulations and experimental particle-image-velocimetry measurements. A mean-preserving input intervention removes velocity fluctuations from selected regions of observed flow histories. Across four neural operators, removing fluctuations from the most energetic 10% of valid observed cells changes forecasts more than equal-area random removal. Because the masks are not matched for removed fluctuation energy, this contrast measures sensitivity, not independent evidence of physical importance. Separately, a CNO has lower velocity-field error but substantially higher two-component fluctuation-energy error than the reference on both analysis subsets. An output attenuation stress test also demonstrates disagreement between benchmark errors and domain-summed fluctuation energy. These single-benchmark results motivate reporting complementary physical diagnostics alongside aggregate prediction scores; they do not establish counterfactual physical correctness.
Comments10 pages, 6 figures. Accepted at the NeurIPS 2026 XAI4Science Workshop, Tiny Paper Track
