Your browser doesn't support javascript.
loading
Switchgear Digitalization-Research Path, Status, and Future Work.
Kastelan, Nediljko; Vujovic, Igor; Krcum, Maja; Assani, Nur.
Afiliação
  • Kastelan N; Faculty of Maritime Studies, University of Split, Ul. Rudera Boskovica 37, 21000 Split, Croatia.
  • Vujovic I; Faculty of Maritime Studies, University of Split, Ul. Rudera Boskovica 37, 21000 Split, Croatia.
  • Krcum M; Faculty of Maritime Studies, University of Split, Ul. Rudera Boskovica 37, 21000 Split, Croatia.
  • Assani N; Faculty of Maritime Studies, University of Split, Ul. Rudera Boskovica 37, 21000 Split, Croatia.
Sensors (Basel) ; 22(20)2022 Oct 18.
Article em En | MEDLINE | ID: mdl-36298270
To keep pace with global energy efficiency trends and, in particular, emission reduction targets in the maritime sector, both onshore and maritime power distribution systems need to be adapted to the relevant new technologies and concepts. As an important link in the distribution chain, medium-voltage switchgear (MV) is expected to be stable and reliable while operating as efficiently as possible. Failures of MV equipment, while rare because the equipment must be safe to handle and use, have far-reaching consequences. The consequences of such failures, whether to the shore or marine power system, present risks to the entire power plant, so an accurate assessment of equipment condition is required to identify potential failures early. The solution is an emerging concept of digital switchgear, where the implementation of sensor technology and communication protocols enables effective condition monitoring, and the creation of a database that, when combined with machine learning algorithms, enables predictive maintenance and/or fault detection. This paper presents, step by step, the previous challenges, the current research (divided into predictive maintenance, condition monitoring, and fault detection categories), and the future directions in this field. The use of artificial intelligence is discussed to eliminate the disadvantage of manually interpreting operational data, and recommendations for future work are formulated, such as the need to standardize test procedures and data sets to train and compare different algorithms before they are used in practice.
Assuntos
Palavras-chave

Texto completo: 1 Base de dados: MEDLINE Assunto principal: Algoritmos / Inteligência Artificial Tipo de estudo: Guideline / Prognostic_studies Idioma: En Ano de publicação: 2022 Tipo de documento: Article

Texto completo: 1 Base de dados: MEDLINE Assunto principal: Algoritmos / Inteligência Artificial Tipo de estudo: Guideline / Prognostic_studies Idioma: En Ano de publicação: 2022 Tipo de documento: Article