Reliability and sensitivity analyses of monopile supported offshore wind turbines based on probability density evolution method with pre-screening of controlling parameters

Abstract

To reasonably investigate the influence of pile-soil interaction on the dynamic reliability of monopile supported offshore wind turbines (OWTs), a framework with high efficiency and good accuracy is proposed for reliability and sensitivity analyses of monopile supported OWTs. This proposed framework combines the grey relational analysis (GRA), the probability density evolution method (PDEM), and the interval mathematical analysis method (IM). Firstly, the GRA is applied to identify the main controlling parameters influencing the stochastic dynamic response of a monopile supported OWT, considering various factors such as pile elasticity modulus, soil internal friction angle and soil unit weight. Secondly, combining the probability density evolution theory, a complete sample probability set is established to obtain the dynamic reliability of the monopile supported OWT under wind and wave loads. Finally, the effect of each controlling parameter on the dynamic reliability of the monopile supported OWT is systematically analyzed by combining the IM and the change of probability measure. The proposed GRA-PDEM-IM-based framework for reliability analyses of monopile supported OWTs can effectively reduce the number of samples for probability density evolution calculations and provide reasonable and accurate reliability analyses results which can serve as a reference for design and construction of monopile supported OWTs.