A comparison between two methods to estimate reliability function for two parameters Extreme value distribution

This paper presents two methods, white and modifies white methods, to estimate reliability function for extreme value distribution as an important model of life time models. The methods enablereliability analysis and prediction in the case of complete samples. The mean square errors of the estimates are computed. Comparisons are made between these estimators using Monte Carlo simulation study, by taking many real values for parameters of extreme value distribution ( θ = 1,6) and (y= 3) using sample sizes (n = 15,25,50) and run sizes (N= 100,500,1 000). The reliability function was estimated at many times to failure (T=l ,3,6). An experiment on bearing simulationtesting was employed to validate the modify white method best C01I1pared with white method.