CasEm：用于长时程神经仿真级联架构
CasEm: A Cascade Architecture for Long-Horizon Neural Emulation
浏览论文内容
中文总结 AI 辅助
CasEm通过独立演化的聚合体模型指导全状态预测修正，在四个基准和气候仿真中显著降低长时程误差，且推理开销极小。
中文摘要 AI 辅助
自回归神经仿真器尽管短期预测准确，但在长时间展开时可能会漂移或发散。我们引入了级联仿真（CasEm），这是一种单向展开架构，通过一个独立演化的物理指定聚合体模型来增强现有的全状态主干网络。其预测指导对全状态预测的修正，而无需从主干网络向聚合体模型反馈。有效的指导要求聚合体能够覆盖显著的主干误差、保持准确可预测性，并支持有用的全状态修正。我们推导了一个有限时域误差界，阐明了这三个因素，并使用经验诊断来指导子系统选择。在四个ODE/PDE基准测试中，CasEm在多种主干网络上减少了长时间展开误差，并在使用傅里叶神经算子主干的两个扩散任务中抑制了误差发散的趋势。在全球气候仿真中，带有区域总水量子系统的CasEm分别将冻结ACE和球形DYffusion主干的10年全状态时间平均误差降低了66.6%和46.3%，同时推理时间增加不到3%。
英文摘要
Autoregressive neural emulators can drift or diverge over long rollouts despite accurate short-term predictions. We introduce Cascaded Emulation (CasEm), a one-way rollout architecture that augments an existing full-state backbone with an independently evolving model of physically specified aggregates. Its forecasts guide corrections to full-state predictions, without feedback from the backbone to the aggregate model. Effective guidance requires aggregates that cover substantial backbone error, remain accurately predictable, and support useful full-state corrections. We derive a finite-horizon error bound that clarifies these three factors and use empirical diagnostics to guide subsystem selection. Across four ODE/PDE benchmarks, CasEm reduces long-horizon rollout errors across diverse backbones and suppresses the trend toward error divergence in both diffusion tasks using Fourier neural operator backbones. In global climate emulation, CasEm with a regional total-water subsystem reduces 10-year full-state time-mean error by 66.6% and 46.3% for frozen ACE and Spherical DYffusion backbones, respectively, while adding less than 3% to inference time.
发表机构
- Southeast University（东南大学）
- Beijing Zhongguancun Academy（北京中关村学院）
- Tsinghua University（清华大学）
机构由 AI 辅助整理，请以论文原文为准。
