A Multi-head Attention Fusion Network for Industrial Prognostics under Discrete Operational Conditions
Yuqi Su 1, Xiaolei Fang 2 · 2026
Read A Multi-head Attention Fusion Network for Industrial Prognostics under Discrete Operational Conditions by Yuqi Su 1, Xiaolei Fang 2 (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2604-10248.txt.
Description
Su and Fang model factory forecasts under shifting work conditions with a multi-head attention fusion network. Engines, turbines, and machines wear in their own ways as work shifts, making forecasts hard. Their method fuses three signal parts: a wear trend, discrete work states from grouping encoded as vectors, and leftover noise. BiLSTM networks capture time links, weighting time steps and sensor signals as needed. A fusion module joins the trend branch with state vectors to capture their links. Tests on a NASA data set confirm the method works.
Catalog record
| CREATOR | Yuqi Su 1, Xiaolei Fang 2 |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 45847 |
| LICENSE | http://creativecommons.org/publicdomain/zero/1.0/ |
| RIGHTS | CC0 |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2604.10248