PERFORMANCE MEASURES AND PARAMETRIC ESTIMATION OF A PARALLEL-SERIES SYSTEM MODEL WITH WEIBULL FAILURE AND REPPAIR TIME DISTRIBUTIONS

The present paper aims at the study of three non-identical unit parallel-series system model with Weibull failure and repair time distributions. The main focus is on the analysis of reliability characteristics and also estimation of parameters in Classical and Bayesian paradigms. The units in the system are named as A, B and C. In order that system works successfully, unit A and anyone of the units B or C should work.Upon failure a unit will immediately be taken up to repair facility for necessary remedial actions. A single repair facility is available to deal with any kind of failure/fault detected with any unit. First come first served service discipline is followed. The failure and repair time distributions for all the units are taken Weibull with varying parameters. Regenerative point technique is used to study various measure of effectiveness. MLE and Bayes estimates of various failure and repair parameters involved in the study have been obtained. A simulation study has also been undertaken to exhibit the behaviour of obtained characteristics in Classical and Bayesian setup and a comparison is made thereupon. Various conclusions have been drawn from the tables and graphs plotted for numerous performance measures for varying values of repair and failure parameters. Having obtained all the reliability characteristics, the profit incurred by the system has been obtained and studied graphically too.