Model predictive control of wind turbine based on deep-dive holistic observer of tower top IMU

In the context of cost reduction and efficiency improvement, with the continuous increase of single unit capacity of wind turbines, their units have become increasingly complex and fragile. The uncertainty of various factors makes wind turbines a highly uncertain nonlinear system, which poses many challenges to the research and control of wind turbines. Therefore, in order to ensure the safe, stable, and efficient operation of wind turbines, this paper proposes a model predictive control (MPC) method based on deep-deep holistic observer (DDHO) of tower top inertial measurement unit (IMU) to maximize power generation, alleviate fatigue loads on the turbines, and extend their service life. Firstly, tower top IMU sensor is used to measure parameters such as wind turbine rotor speed and acceleration. Then, by using DDHO combined with measurement parameters, the wind speed and unknown state of the wind turbine model are estimated. Then, by reducing phase lag in state estimation and adding dampers, the damping effect of the tower can be enhanced. Finally, the estimated parameters are used as inputs for MPC to perform real-time prediction processing on the predicted time-domain state values, in order to obtain the optimal control method for the system at the current time by solving the minimum objective function, minimizing the difference between the system reference trajectory and future output values. Through experiments of 10 MW wind turbine, it has been proven that the proposed control method can improve the power generation efficiency of wind turbines to a certain extent, reduce actuator wear, and reduce tower fore-aft fatigue loads.