Naslov
Determining the frequency of condition based maintenance tasks with neural network

Autori
Šarić, Tomislav ; Majdandžić, Niko

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
1st DAAAMInternationalConference on Advanced Technologies for Developing Countries / Katalinić, Branko - Slavonski Brod, 2002, 533 - 539

Skup
1st DAAAMInternationalConference on Advanced Technologies for Developing Countries

Mjesto i datum
Slavonski Brod, Hrvatska, 12.09.2002. - 14.09.2002

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Condition Based Maintenance; Neural Networks; Bering

Sažetak
Use of modern maintenance approaches and strategies in production process ensures a high degree of machine and equipment availability and reliability to companies. Among other factors, this imposes itself as an important assumption in realization of production plan and dynamic market demands. Use of modern strategies means an implementation of knowledge, equipment and organization. Proper implementation declares itself in a cost-efficient approach. /1/. Condition based maintenance, in general, can be explained through several characteristic steps it is made up of, and those are: -Technical diagnostic, which, among other, determines the condition-indicators of the observed technical elements ; -Using historical data, we can analyse the changes in the trend of the indicator’ s condition in time ; -Finally, we can define the problem on the level of prediction of the element’ s condition in future.

Izvorni jezik
Engleski

Znanstvena područja
Strojarstvo

POVEZANOST RADA