AI4Plasma I：五种神经网络架构用于容性耦合等离子体场代理建模的数据效率对比
AI4Plasma I: Data-Efficient Surrogate Modeling of Capacitively Coupled Plasma Fields
做物理仿真代理建模的可以看这篇，它把五个主流架构在等离子体仿真上的样本效率拉通了对比，FNO 多数场景最省数据，但边界工况下 Transolver 反超。选架构别只看榜单，要看你的精度目标和参数区间。
arXiv 论文 AI4Plasma I 对比了五种代理模型架构——Coordinate MLP、PCA-Net、U-Net、FNO 和 Transolver——在容性耦合等离子体（CCP）仿真中预测 RF 周期平均电子密度、离子密度、电势和电子温度所需的最少参考算例数。
核心发现是 FNO 在八个对比项中的七项达到最低或并列最低的数据需求：1D 情形下 FNO 用 50 个训练算例即可达到 1% 内部误差，Transolver 需要 100 个，U-Net 需要 200 个；2D 情形下 FNO 和 U-Net 各用 100 个算例达到 1% 内部误差，但在更严格的 2D 1% 参数边界目标下，Transolver 只需 400 个算例，FNO 反而需要 800 个。
论文还在 1D 平面和 2D 轴对称 GEC CCP 基准上分析了 FNO 谱模式与填充、U-Net 跳跃连接、Transolver 自适应切片及 PCA-Net 投影误差等机制，结论是 FNO 可作为 CCP 场景的数据高效基线，但最优架构取决于精度目标与工况参数区间。
AI4Plasma I: Data-Efficient Surrogate Modeling of Capacitively Coupled Plasma Fields
Reference capacitively coupled plasma (CCP) simulations provide spatially resolved plasma fields but are costly to generate across operating conditions. We compare the data efficiency of five surrogate architectures--Coordinate MLP, PCA-Net, U-Net, Fourier neural operator (FNO), and Transolver--for predicting RF-cycle-averaged electron density, ion density, electric potential, and electron temperature in 1D planar and 2D axisymmetric GEC CCP benchmarks. We assess the number of reference simulations needed to achieve 1% and 2% mean relative L2 error for parameter-interior and parameter-boundary test cases, and examine architecture-specific mechanisms. FNO achieves the lowest or joint-lowest reference-case requirement in seven of eight comparisons. In 1D, FNO reaches 1% interior error with 50 training cases, compared with 100 for Transolver and 200 for U-Net. In 2D, FNO and U-Net each reach 1% interior error with 100 cases; FNO and Transolver each reach 2% parameter-boundary error with 200 cases, compared with 400 for U-Net. At the stricter 1% 2D parameter-boundary target, Transolver requires 400 cases versus 800 for FNO. Additional analyses investigate FNO spectral modes and padding, U-Net skip connections, Transolver adaptive slicing, and PCA-Net projection versus prediction error. The results support FNO as a data-efficient baseline for the tested CCP settings, while showing that the preferred architecture depends on the accuracy target and operating-parameter region.
本条情报归属 AITOP「论文」核心追踪栏目。在跨区域产业观察坐标中，该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线，反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括：FNO、Transolver、神经网络算子、代理建模，读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明：本文原始线索采编自「arXiv cs.LG」（@Zhenyu Zou, Junjue Liao, Hao Yang, Luowei Yin, Chenyu Wang, Yufeng Wei, Zhichao Peng, Chun Chen, Daniel Zhengyu Huang, Kangxin Liu, Haoran Guan, Rubing Han, Yipei Chen, Peiting You, Yuwei Fan），经 AITOP 多维评分管线结构化提炼。