A neural network approach for machine breakdown repair time

摘要

Research on neural network applications have been carried out very extensively inudrecent days. The current trends in manufacturing sectors for solving their businessudoperational problems have been very difficult and subjective. Many organizationsudhave used various methods to solve machine breakdown's repair time, either reducingudthe time taken to repair or eliminate the particular occurrence. The traditional wayudfor solving these machine breakdown issues was to predict the machine breakdownudoccurrence through preventive maintenance. Hence, in the present study, a neuraludnetwork method was proposed to optimize the mean repair time for machineudbreakdown with regression models were evaluated from the trained neurons. Theudneurons were represented by the samples of repair time of previous years' record of audsingle machine. The results shows that the set of samples of repair time haveudcritically influenced the optimized mean repair time for the machine. Variousudmethodologies were used by comparing several grouped machine breakdownudphenomena which showed more accurate regressions. The use of neural network, inudthe end of the study, gives significant changes in predicting machine breakdownudrepair time for the future years.