Reproduction code and derived degradation trajectories for indicator-aware design of run-to-failure degradation tests

Résumé fourni par la source

Reproduction code for the manuscript "Indicator-aware design of run-to-failure degradation tests for data-driven prognostics" by Huy Hoang Le and Kim-Anh Nguyen. Every measured quantity, table and figure in the manuscript is produced by one of the 108 scripts in experiments/. README.md maps each table and each figure to the script that produces it. The archive also contains the derived arrays behind the battery and bearing results — per-cell capacity and Delta-Q(V) trajectories, cycle lives, extracted features, bootstrap intervals, and cached bearing and PRONOSTIA features — so that those analyses, and all four figures of the manuscript, run with nothing downloaded. The raw run-to-failure records are too large to redistribute and are obtained from their public sources; set_paths.py points the code at them. Datasets analysed: Severson et al. lithium-ion cells; C-MAPSS FD001 and FD003 turbofans; XJTU-SY and PRONOSTIA/FEMTO-ST bearings; NASA Ames milling; NASA PCoE batteries; Zhang et al. impedance cells; TJU cells. All are publicly available and are cited in the manuscript. Derived arrays are redistributed under the terms of their source datasets, with attribution given in README.md. Requires Python 3.11 or later with numpy, scipy, scikit-learn, matplotlib and sympy; see requirements.txt.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.