Bayes Reliability Estimation of Motor Drives Redundant Configuration

Motor drives reliability is a crucial problem for power system reliability, availability and maintainability, and is being the object of more and more significant studies. In many industrial, aerospace and naval applications is necessary to use parallel configuration of electrical drives, with the aim to split the power into more electrical drives and obtaining reserve units. For a typical structure based on the use of “n” parallel electrical drives with “(n-k)” reserve units, the present study develops a novel Bayesian statistical inference approach for the estimation of such typical redundant system, which is defined as a “k out of n” redundant reliability configuration, being k and n positive integer numbers, with 1≤k≤n. The proposed approach is based upon the classical Exponential reliability for each component and uses typical available data for the hazard rate of the components, which is characterized through a Gamma prior distribution. Extensive numerical simulations confirm that the proposed estimation technique constitutes a very simple, and efficient method for the above system reliability estimation, both in terms of point estimation and interval estimation, performed through a suitable Beta distribution. Also, some hints to a robustness analysis with respect to other different prior distributions are reported in the study.