自监督预训练用于 PHM：什么能迁移、什么不能、联邦化代价几何
Self-Supervised Pretraining for PHM: What Transfers, What Does Not, and What Federating It Costs
AI 导读
一项研究在 47 个公开 PHM 数据集上、以防泄漏协议测量自监督预训练对故障预测与健康管理（PHM）的价值。用掩码补丁预测预训练的 801,808 参数通道独立 Transformer 编码器，将 CWRU 故障分类 macro-F1 从 0.350 提至 0.812、HSG18 从 0.569 提至 0.951，但对 CMAPSS 剩余寿命（RUL）估计无可靠提升。
正文 · 原文
Abstract:Self-supervised pretraining has produced capable time-series foundation models, but its value for Prognostics and Health Management (PHM) has not been measured under leakage-controlled conditions. We measure it on a corpus of 47 public PHM datasets, harmonized to a single tensor contract and partitioned so that 11 datasets are withheld from pretraining in full. Transfer turns out to be task-dependent: a compact channel-independent Transformer encoder (801,808 parameters) pretrained with masked patch prediction raises fault-classification macro-F1 from 0.350 to 0.812 on CWRU and from 0.569 to 0.951 on HSG18, yet on CMAPSS remaining-useful-life estimation no pretrained checkpoint reliably improves upon a from-scratch baseline. We pretrain under two regimes---centralized, and simulated cross-silo federated with ten clients grouped by physical domain---holding architecture, objective, source data and optimization budget fixed. Two further findings are negative. At matched budget the centralized encoder leads by 15.5--20.2 macro-F1 points in four of four comparisons, although the comparison is confounded by a learning-rate schedule that favours the centralized arm and rests on a single pretraining run per regime; on RUL the two regimes are indistinguishable. And the intrinsic SSL loss is not a valid selection criterion: the centralized encoder more than halves the federated reconstruction loss without a corresponding gap on RUL, and a server-momentum sweep moves downstream validation macro-F1 by up to 0.29 while the SSL loss varies by 0.021. Transfer also hinges on the backbone learning rate: the same checkpoint yields 0.169 or 0.829 macro-F1 on CWRU depending on a single order of magnitude. We release the corpus pipeline, the protocol and the code.
| Subjects: | Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Machine Learning (cs.LG); Performance (cs.PF) |
| Cite as: | arXiv:2610.35280 [cs.AI] |
| (or arXiv:2610.35280v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.35280 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: David Solís-Martín [view email]
[v1]
Mon, 26 Oct 2026 11:50:33 UTC (937 KB)
来源：arXiv cs.AI · arxiv.org