Estimation of a repairable system reliability

Abstract

Maintaining production equipment is a key issue, ensuring optimum productivity and excellent quality of products or services. It is an industrial challenge involving the questioning of today's fixed structures and the promotion of methods adapted to the new nature of materials. The economic performance of a production entity is dependent on the reliability of its equipment, therefore the control of the maintenance policy. To meet these requirements, this paper proposes a structured methodology based on powerful probability laws that describe the behavior of an equipment; recent statistics laws to evaluate the parameters. The SVR (Support Vector Regression) method is used to identify the parameters of Weibull law of; consequently, the degradation law control of a good allows to react at the opportune moment.