Szczegóły publikacji
Opis bibliograficzny
Bayesian optimal sensor placement for acoustic emission source localization with clusters of sensors in isotropic plates / Siddhesh RAORANE, Tulay Ercan, Costas Papadimitriou, Paweł PAĆKO, Tadeusz UHL // Mechanical Systems and Signal Processing ; ISSN 0888-3270. — 2024 — vol. 214 art. no. 111342, s. 1-21. — Bibliogr. s. 20-21, Abstr. — Publikacja dostępna online od: 2024-03-28
Autorzy (5)
- AGHRaorane Siddhesh
- Ercan Tulay
- Papadimitriou Costas
- AGHPaćko Paweł
- AGHUhl Tadeusz
Słowa kluczowe
Dane bibliometryczne
| ID BaDAP | 152622 |
|---|---|
| Data dodania do BaDAP | 2024-04-24 |
| Tekst źródłowy | URL |
| DOI | 10.1016/j.ymssp.2024.111342 |
| Rok publikacji | 2024 |
| Typ publikacji | artykuł w czasopiśmie |
| Otwarty dostęp |  |
| Czasopismo/seria | Mechanical Systems and Signal Processing |
Abstract
For practical applications of acoustic emission (AE) source localization, it is extremely important to optimally place the sensors, i.e., the sensors should be placed such that they gather the most effective information — leading to highly accurate localization of AE sources. In this paper, a Bayesian optimal sensor placement strategy is used to determine the optimal (and worst) positions of sensor clusters for AE source localization in isotropic plates with unknown material properties. The sensor clusters are composed of three sensors arranged in a right-angle triangular configuration, and the AE source localization strategy employed requires placement of at least 2 sensor clusters. In the work presented here, three different hot-spot (source location) areas are analyzed, and the best and worst clusters positions are predicted for the placement of 2, 3 and 4 sensors clusters. The optimization required to arrive at the optimal positions is performed using exhaustive search method, heuristic search methods and genetic algorithms. Furthermore, theoretical validation - using Bayesian inference with simulated data - and experimental validation - using AE source localization methodology and Bayesian inference with experimental data - are presented to validate the predictions by the Bayesian optimal sensor placement.
