Naslov
Decision Support for the Fault Diagnossis-Possibility of Deep Drawing Reengineering

Autori
Ćosić, Predrag

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

Izvornik
CARS & FOF 2001 / Glen, Bright & Wendy, Janssens - Durban : The Mechatronics and Robotics Research Group, Univ. of Natal, 2001, 137-143

Skup
17th International Conference on CAD/CAM, Robotics and Factories of the Future

Mjesto i datum
Durban, Južnoafrička Republika, 10.07.2001. - 12.07.2001

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
deep drawing; fault diagnosis; fuzzy variables; certainty factor algebra; reengineering

Sažetak
The development of an off-line expert system prototype dedicated to quality defect diagnosis in deep drawing technology is described. Database creating, database search strategy, defining of logical relationships and reasoning system reliability are outlined. The diagnosis of an observed defect or failure, applying a set of rules, gives the expert system user an insight into the potential classified and graded faults. After the multilevel validation of potential causes of the observed faults, the expert system chooses the solution which is based on the greatest probability of the evaluation function. In this paper the use of fuzzy system modeling in expert system for the purpose of detection and elimination of faults in the technology of deep drawing is considered. The evaluation function estimates the possible causes of the faults and suggests actions taking into account the following criteria: the product price, time of use, level of disturbing technological process, and standardisation of tool parts and technological parameters.

Izvorni jezik
Engleski

POVEZANOST RADA