AI 中文总结
针对异构IoT传感器多变量时间序列预测问题，提出CrossRAG框架，结合SAM、FCC学习与CATF，在7个基准上性能优于仅参数化基线及现有检索增强预测方法。
AI 中文摘要
来自异构物联网传感器的大规模多变量时间序列，为资源调度和预测性维护提供了精准长期预测的需求。近期时间序列基础模型虽展现出强大的泛化能力，但依赖静态参数化知识，推理过程中缺乏对外部历史模式的动态访问。检索增强生成（RAG）或可解决该问题，但其在时间序列预测中的应用面临异构源间幅值差异、历史相似性与未来一致性不匹配的挑战。本文提出CrossRAG，一种检索增强的预测框架，整合了形状感知记忆（SAM）与RevIN归一化以实现幅值鲁棒的形状级检索，未来一致对比（FCC）学习以区分信息丰富的参考与历史相似但未来分歧的难负样本，以及交叉注意时间融合（CATF）以在表示层面将检索到的历史-未来参考对融合到骨干网络的表示中。在7个公共基准上的实验表明，CrossRAG始终优于仅参数化基线和现有检索增强预测方法。
英文摘要
Large-scale multivariate time series from heterogeneous IoT sensors demand accurate long-term forecasting for resource scheduling and predictive maintenance. While recent time series foundation models exhibit strong generalization, they rely on static parametric knowledge and lack dynamic access to external historical patterns during inference. Retrieval-Augmented Generation (RAG) offers a potential remedy, yet its application to time series forecasting is challenged by magnitude variations across heterogeneous sources and the mismatch between historical similarity and future consistency. We propose CrossRAG, a retrieval-augmented forecasting framework that integrates Shape-Aware Memory (SAM) with RevIN normalization for magnitude-robust shape-level retrieval, Future-Consistent Contrastive (FCC) learning to distinguish informative references from hard negatives with similar history but divergent futures, and Cross-Attention Temporal Fusion (CATF) to fuse retrieved historical--future reference pairs into the backbone's representations at the representation level. Experiments on seven public benchmarks show that CrossRAG consistently outperforms both parametric-only baselines and existing retrieval-augmented forecasting methods.
