Title :
A statistical method for evaluating electrical failures
Author :
Barrett, J.S. ; Green, M.A.
Author_Institution :
Ontario Hydro Technol., Toronto, Ont., Canada
fDate :
7/1/1994 12:00:00 AM
Abstract :
Statistical tools have been developed for performing Bayesian analysis of electrical failures in the case of two-parameter distributions. Software implementations graphically present confidence regions for the two parameters and confidence intervals for the failure probability as a function of applied stresses. For the normal and lognormal distributions these are the same as classical confidence regions, while for Gumbel and Weibull distributions they are nearly the same. Examples are provided for the energy capacity of metal-oxide varistors and for insulator flashover probability
Keywords :
Bayes methods; failure analysis; flashover; insulation testing; insulators; probability; surge protection; varistors; Bayesian analysis; Gumbel distribution; Weibull distribution; applied stresses; confidence intervals; confidence regions; electrical failures evaluation; energy capacity; failure probability; insulator flashover probability; lognormal distribution; metal-oxide varistors; normal distribution; software implementations; statistical method; two-parameter distributions; Bayesian methods; Capacity planning; Failure analysis; Metal-insulator structures; Performance analysis; Probability; Statistical analysis; Stress; Varistors; Weibull distribution;
Journal_Title :
Power Delivery, IEEE Transactions on
