Estimation of Reliability through the Wiener Degradation Process Based on the Genetic Algorithm to Estimating Parameters

Abstract

In this paper, the researcher suggested using the Genetic algorithm method to estimate the parameters of the Wiener degradation process, where it is based on the Wiener process in order to estimate the reliability of high-efficiency products, due to the difficulty of estimating the reliability of them using traditional techniques that depend only on the failure times of products. Monte Carlo simulation has been applied for the purpose of proving the efficiency of the proposed method in estimating parameters; it was compared with the method of the maximum likelihood estimation. The results were that the Genetic algorithm method is the best based on the AMSE comparison criterion, then the reliability was estimated by an inverse Gaussian distribution according to the characteristics of the Wiener process. It was also applied based on real data taken from an experiment intended to determine the degradation in the light intensity of lamps under specific experimental conditions