Lidar Aided Feedforward Control Of Large Wind Turbine

With the increasing global demand for wind energy utilization,wind turbine is developing towards large capacity and large-scale.In order to improve the power generation capacity of wind turbine and reduce its load,advanced measurement technology and control technology are needed to achieve this goal.The rapid development of lidar wind measurement technology provides an application basis for feedforward control to serve wind turbine control.However,the information measured by lidar is the line of sight wind speed at a distance in front of the wind turbine impeller,which can not be directly used in the feedforward control of wind turbine.Therefore,this paper focuses on how to apply lidar wind measurement information to wind turbine feedforward control.The main work includes:(1)In order to accurately predict the equivalent wind speed on the blade surface,a hybrid model based on empirical mode decomposition and gated cyclic element neural network is proposed.Firstly,the wind speed information measured by lidar is decomposed by empirical mode to reduce the interference of mechanism model parameters on the prediction accuracy of equivalent wind speed on the impeller surface,then each decomposed connotative mode component is predicted,and finally the final equivalent wind speed on the impeller surface is obtained by aggregation calculation.In order to verify the effectiveness of the proposed method,the simulation verification is carried out based on bladed software,and the equivalent wind speed prediction results obtained by using the model are compared with the equivalent wind speed prediction results based on mechanism modeling and the prediction results based on gated cycle unit neural network model.The simulation results show that this method can accurately and reliably predict the equivalent wind speed on the impeller surface.(2)Based on the feedforward control scheme commonly used in industry,a new feedforward control scheme based on dynamic nonlinear compensation is proposed.In order to deal with the uncertainty of lidar wind measurement,the control objectives of decomposition feedforward controller are rotor speed and rotor acceleration;According to the rotor acceleration control target,the inverse model is constructed by using the flatness property of the fan simplified model to obtain the feedforward compensation of the corresponding control input.Simulation experiments and comparative verification of the control strategy are carried out based on the two cases.The simulation results show that the proposed feedforward control scheme based on dynamic nonlinear compensation has better performance than the reference controller in all aspects under extreme and conventional conditions.(3)Considering the uncertainty of equivalent wind speed prediction,a feedforward control of wind turbine based on multi scenario optimization is further proposed.In order to reduce the impact of inaccurate wind speed prediction,the predicted equivalent wind speed is compared with the real equivalent wind speed based on the simulation experimental data,and the probability distribution of wind speed prediction error and the change degree of equivalent wind speed are obtained;Then,it is discretized into multiple scenes to describe the real wind speed information in the future;Finally,the feedforward control scheme is designed according to the needs of multiple scenes,and the multi scene feedforward control framework based on fuzzy weight coefficient adjustment is established,and the simulation experiment is carried out to verify it.The simulation results show that this method can deal with the influence of prediction error,and the control performance is better than the single feedforward control.