A dynamic probabilistic analysis method for wind turbine rotor based on the surrogate model

Abstract

Due to the long and flexible characteristics of the wind turbine blade, the influence of fluid-structure interaction on the dynamic response results cannot be ignored. The dynamic analysis of the wind turbine rotor based on the fluid-structure interaction is very computationally expensive, and the dynamic reliability analysis considering the influence of randomness usually requires a large number of computational samples. In this study, a dynamic probabilistic analysis method based on the surrogate model (SM-DPAM) to the characteristics of the wind turbine rotor was established, combining numerical simulation, intelligent algorithms, and data mining methods for wind turbines. This method allows for fast and inexpensive reliability and sensitivity analysis by building accurate surrogate models with a limited number of expensive fluid-structure interaction (FSI) samples. In a case of 5MW wind turbine rotor, the average relative error of the test was 0.093%, the reliability was 0.9515 and two variables insensitive variables were found. It is shown that the method can effectively analyze the reliability and sensitivity of wind turbine rotors, adapted to the nonlinear and highly dimensional characteristics of wind turbines well, and reduces the cost of wind turbine studies by controlling the number of expensive samples used. In addition, this research provides a reference for using intelligent algorithms and data mining methods in wind turbine design.