Improved Estimation of Weibull Parameters with the Linear Regression Method

Abstract

Monte Carlo simulations were used to search for the optimal probability estimator for estimating Weibull parameters with the linear regression method. Compared with commonly used probability estimators, the optimal one obtained gives a more accurate estimation of the Weibull modulus and the same estimation precision of the scale parameter. It is also concluded that the maximum likelihood method results in the highest precision, however, less conservative than the linear regression method.