Research On Power Optimization Method For Large-Scale Wind Turbine Based On Adaptive Model Predictive Control

The characteristics of large-scale wind turbines,such as parameter uncertainties,strong nonlinearity,and large rotational inertia,pose challenges in control due to the wide range of wind speed variations during operation.In response to the issues of model mismatch and degraded control performance associated with control methods based on single equilibrium point linearization,this study integrates effective wind speed estimation and a linear parameter varying model of the wind turbine to establish an adaptive predictive control framework.The proposed approach adapts and updates the predictive model parameters within the control time interval,aiming to optimize power generation by achieving maximum power point tracking below rated wind speed and enhancing output power stability above rated wind speed.Simulation studies are conducted using the Bladed software with the incorporation of an external controller.The research outcomes presented herein are supported by the National Key Research and Development Program project "Optimization Design,Batch Manufacturing Processes,and Inspection Techniques for Key Components in Large-scale Offshore Wind Turbine Drive Systems"(2018YFB1501304).The primary research work and achievements are as follows:(1)The operational characteristics and control mechanisms of wind turbines were investigated,and a mathematical model of the wind turbine was established based on the structural dynamic characteristics of its subsystems.In response to the large rotational inertia exhibited by large-scale wind turbines,a proportional-integral control-based method for maximum power point tracking using the tip speed ratio was proposed,incorporating effective wind speed estimation.(2)The fundamental principles of adaptive model predictive control were analyzed.In response to the model mismatch issue caused by the uncertainties and nonlinear characteristics in the operation of wind turbines,the dynamic model of the wind turbine was linearized and discretized at multiple operating points,resulting in a linear parameter-varying state-space model.By integrating an effective wind speed estimation method based on the Kalman filtering algorithm,a linear parameter-varying adaptive predictive model was obtained.(3)The adaptive model predictive control power optimization method based on the linear parameter-varying adaptive predictive model was proposed,and the traditional model predictive control power optimization method was introduced as a reference.Simulation results demonstrated that in the region below the rated wind speed,the adaptive model predictive control achieved a power increase of 1.4%and 0.6%compared to the optimal torque control and model predictive control,respectively,effectively achieving maximum power point tracking.In the region above the rated wind speed,the adaptive model predictive control enhanced the system’s robustness.Compared to the model predictive control,it reduced the standard deviation of the output power by 23.7%under turbulent wind conditions,thus improving the stability of the output power.