增强控制的自回归扩散用于数据同化
Control-Augmented Autoregressive Diffusion for Data Assimilation
AI总结：
本文提出一种增强控制的自回归扩散模型，通过引入预训练模型与离线训练控制器，提升数据同化中混沌时空偏微分方程的稳定性和准确性。
摘要：
尽管在测试时扩展和扩散微调方面有所进展，但对自回归扩散模型（ARDMs）的指导仍显不足。我们介绍了一个可 amortized 的框架，该框架通过将预训练的 ARDMS 与离线训练的控制器相结合，通过预览未来的滚动预测，控制器学习逐步修正，以在终端成本目标下预测观测，从而生成可重用的策略。受 ARDM 轨迹的随机最优控制观点启发，我们的方法在每个去噪子步骤中注入小控制量，同时保持接近预训练的动力学。我们研究了这种方法在混沌时空偏微分方程（PDEs）中的数据同化（DA）应用，其中现有方法往往计算成本高且在稀疏观测下易受预测漂移影响。在推理时，DA 变成一个具有实时修正的前馈滚动预测，相较于强大的扩散基线方法，实现了数量级的加速。在两个典型 PDEs 和一个涵盖六个观测领域的紧凑 ECMWF 再分析 v5（ERA5）试点研究中，我们的方法在稳定性和准确性上均优于现有最先进方法，且在更大规模的 GenCast 研究中也观察到相似的改进。
英文摘要：
Despite advances in test-time scaling and diffusion finetuning, guidance for Auto-Regressive Diffusion Models (ARDMs) remains underexplored. We introduce an amortized framework that augments a pretrained ARDM with an offline-trained controller. By previewing future rollouts, the controller learns stepwise corrections that anticipate observations under a terminal-cost objective, yielding a reusable policy for guided generation. Motivated by a stochastic optimal control view of ARDM trajectories, our method injects small controls within each denoising sub-step while staying close to the pretrained dynamics. We study this approach for dataassimilation (DA) in chaotic spatiotemporal partial differential equations (PDEs), where existing methods are often computationally expensive and susceptible to forecast drift under sparse observations. At inference, DA becomes a feed-forward rollout with on-the-fly corrections, achieving an order-of-magnitude speedup over strong diffusion-based baselines. Across two canonical PDEs and a compact ECMWF Reanalysis v5 (ERA5) pilot spanning six observation regimes, our method consistently improves stability and accuracy over state-of-the-art alternatives, with similar improvements observed in a larger-scale GenCast study.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
