The influence of the pitch angle and tip speed ratio on wind speed evolution in the induction zone of the wind turbine incoming flow is demonstrated through fluid structure interaction and computational fluid dynamics simulation respectively,based on both the original NREL 5 MW wind turbine model and its simplified one.Combined with the theoretical model of wind speed evolution and thrust coefficient,it can be obtained that changing the induction factor in the thrust model and then updating the thrust coefficient in the wind speed evolution model online can improve the accuracy of the effective wind speed estimation.Considering the time-varying and non-stationary nature of thrust,a thrust approximation method based on RBF neural networks with the input of pitch angle and tip speed ratio is put forward.Then,an estimation method of wind turbine plane effective wind speed based on online model updating is proposed.At last,validation is conducted using on-site wind measurement data from LiDAR,and the results show that the wind speed estimation method based on RBF neural networks and model online update improves the estimation accuracy by 15.5%compared to the estimation method with time information extraction algorithm and incoming flow wind speed evolution model.
维普
万方数据