Szczegóły publikacji
Opis bibliograficzny
Non-clustering method for automatic selection of machine operational states / Adam JABŁOŃSKI, Tomasz BARSZCZ, Piotr Wiciak // W: Proceedings of CMMNO 2013 [Dokument elektroniczny] : international conference on Condition Monitoring of Machinery in Non-Stationary Operations : Ferrara, 8 to 10 May, 2013 / eds. Giorgio Dalpiaz [et al.]. — Wersja do Windows. — Dane tekstowe. — [Ferrara : s. n.], [2013]. — Dysk Flash. — S. 1–9. — Wymagania systemowe: Adobe Reader. — Bibliogr. s. 9, Abstr. — Tytuł przejęto ze s. tyt.
Autorzy (3)
- AGHJabłoński Adam
- AGHBarszcz Tomasz
- Wiciak Piotr
Dane bibliometryczne
| ID BaDAP | 73632 |
|---|---|
| Data dodania do BaDAP | 2013-05-17 |
| Rok publikacji | 2013 |
| Typ publikacji | materiały konferencyjne (aut.) |
| Otwarty dostęp |  |
| Konferencja | Third international conference on Condition Monitoring of Machinery in Non-Stationary Operations |
Abstract
A reliable evaluation of technical condition of machinery working under non-stationary conditions requires a rigorous tracking of operational parameters. Therefore, modern condition monitoring systems (CMS) enable reading and registering of process parameters (e. g. speed, load, pressure, etc.) in parallel with acquisition of vibroacoustic signals. Although few tries have been undertaken to develop state-free analysis of vibration signals, currently installed systems still do rely on state-preclassified data. The paper shows how the process, referential data might be automatically transformed into proposition of optimal machine operational states in terms of their number and their range. As indicated by the title, the paper shows common pitfalls coming from implementation of popular clustering approach. The proposed algorithm illustrates is verified on real data from a pitch-controlled wind turbine.
