Naslov
Intelligent predictive decision support system for maintenance

Autori
Šarić, Tomislav ; Majdandžić, Niko ; Lujić, Roberto ; Šimunović, Goran ; Matičević, Gordana

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

Izvornik
Proceesings of the 9th International Scientific Conference on Production Engineering (CIM 2003) : Computer Integrated Manufacturing and High Speed Machining / Cebalo, Roko ; Shulz, Herbert - Zagreb : Hrvatska udruga proizvodnog strojarstva (HUPS), 2003, 439-444

ISBN
953-97181-5-5

Skup
International Scientific Conference on Production Engineering (9 ; 2003)

Mjesto i datum
Lumbarda, Hrvatska, 05.06.2003. - 06.06.2003

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
maintenance; Intelligent systems; neural networks; Maintenance Information System
(maintenance; intelligent systems; Neural networks; Maintenance Information System)

Sažetak
The Condition Based Maintenance (CBM) planning and intervals definition is very complex problem. So, through the understanding of words planning and condition based maintenance approaches result as necessity for providing of input plan parameters, prediction of activity length and dynamic plan generating. Condition Based Maintenance is strategy that different authors defined differently, some as form of preventive maintenance, some as preventive maintenance and some as substantive strategy. It is well known that Condition Based Maintenance consists of: historical data state changing analysis of the elements, condition monitoring and technical state prediction of maintenance elements. Analysis of historical data is focused on exploration of technical systems characteristics in exploitation process and forming of different mathematical models that have to describe these changes in time. Historical data is defined as a trend where observed physical parameters will change their state. These changes can be described with different mathematical functions like: linear, exponential, logarithmic, hyperbolic etc. Different mathematical models try to describe and approximate state changes, but usually they give approximate exact results. Which model will be used depends on complexity of technical system. Technical state prediction of maintenance elements in the future is complex and complicated problem. Activities that have to be taken into consideration commonly with set of various mathematical, statistical and other models for technical state prediction of maintenance elements present approach called Predictive Maintenance. Condition Based Maintenance can be separated into two approaches: with parameters control and control of reliability level. Condition Based Maintenance with parameters control is approach where states of selected physical parameters are monitored occasionally or continually.

Izvorni jezik
Engleski

Znanstvena područja
Strojarstvo

POVEZANOST RADA