|
Organizers |
Reliability Modelling of Aluminium Melting Furnace System
by
S. K. Majumdar
Indian Statistical Institute, Kolkata
Graphical plots and test results for trend of successive failure times of an aluminium melting furnace, a repairable system, indicated that its failure pattern was fit to be modelled by an NHPP stochastic point process model. Tests for reduction of the general NHPP peril rate function form to log-linear, log-quadratic and Weibull process models revealed that any one of the three could be chosen. Hence the parameters of the above three peril rate functions were estimated [ using mle ]. Next adequacy of the three models were tested using model-specific adequacy tests, which revealed that all the three were adequate. To get at the most appropriate NHPP peril rate model form, Radon Nikodym Derivative was gainfully used to discriminate among the three models. Weibull process model turned out to be the most appropriate to characterize the NHPP peril rate function of the given system. Using this model, the reliability function was found and system reliability for additional hours [s,s+t] of operation were estimated.
Date received: August 14, 2001
Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cagd-40.