Reliability Estimation Of A Horizontal Axis Wind Turbine Using Tail Modeling

Abstract

Reliability of highly safe structural systems can be estimated efficiently by using tail modeling. The main idea of tail modeling is to perform a relatively small number of limit-state function calculations through a sampling scheme, specify a threshold value that defines the tail part, and then fit a probability model to the tail part. The selected threshold value has a significant effect on reliability estimations. This paper aims at drawing some guidelines for proper selection of the threshold value using benchmark mathematical example problems with varying number of random variables, level of nonlinearity and level of safety. Finally, tail modeling is applied to reliability prediction of a horizontal axis wind turbine. It is found that the tail modeling can predict the high reliability of horizontal axis wind turbine efficiently and accurately.