Reliability Analysis of Wind Turbine Main Shaft Bearings Based on Active Learning Kriging Algorithm

As a key component of the wind turbine, the reliability level of the main shaft bearing seriously affects the performance and safety of the wind turbine. Therefore, it is very important to accurately analyze the reliability of its spindle bearing. By analyzing a certain type of wind turbine cylindrical roller bearing, we consider its complex failure mechanism and the characteristics of implicit function and high nonlinearity. The reliability analysis method of wind turbine spindle bearing based on PAK-B<sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">n</sup> algorithm is proposed. A three-dimensional model of the bearing assembly has been developed, followed by a comprehensive static structural analysis to evaluate its performance under applied loads. The corresponding response values are obtained by changing the input random variables, which are used as a set of data to construct and iterate the Kriging surrogate model. Compared with the Monte Carlo simulation method (MCS), the number of iterations, failure probability and relative error are compared and analyzed. The results show that the failure probability of the proposed method is 0.0022, the reliability is 0.0978, the number of iterations is 72, and the relative error of the failure probability is 1.84 %. Therefore, the reliability analysis method of wind turbine spindle bearing based on active learning Kriging algorithm has high calculation efficiency under the condition of ensuring a certain accuracy. It can be referred by decision makers in practical engineering.