New analytical method for estimating mean life of electric power equipment based on complete and right-censored failure data

- A method to estimate Weibull distribution parameters and lifetime is presented.
- The method is analytical and valid for complete and censored data.
- The solution to the likelihood equation is obtained using asymptotic methods.
- The method is highly precise and making possible error estimation and control.
- The main result of the proposed estimation method is that provides closed-form expressions.

Analysis based on application of two-parametric Weibull distribution is commonly used to characterize power equipment life distributions in the presence of censored data. Numerous studies have focused on enhancing the accuracy of statistical parameter estimation. However, there is still no straightforward and rigorous analytical method of estimation of the Weibull parameters and the mean life. In this work, a new method for mean life evaluation of power system equipment, based on the two-parameter Weibull distribution, is presented for complete and right-censored failure data. Classical maximum likelihood estimation (MLE) is employed for determination of distribution parameters. However, an estimator based on asymptotic expansions is proposed, overpassing the disadvantages of MLE-based methods that employ numerical or graphical techniques. High accuracy of the proposed method with respect to other estimators is also shown by analyzing two right-censored lifetime data sets with different sample sizes for three types of power equipment. The efficiency and accuracy of the proposed analytical method find their strengths in the analytically obtained closed-form expressions for the distribution parameters.