Neural network adaptive control for constant output power of variable pitch wind turbine

This paper proposes a control strategy based on neural network to regulate the blade pitch angle and rotate speed of the wind turbine system. An affine model of wind turbine system is proposed. On the one hand, from the aerodynamic model one can know that there exists strong coupling between the wind turbine speed and the pitch angle. On the other hand, the range of pitch angle and variable pitch rate of the wind turbine system is limited. Therefore, the wind turbine system should be seen as a nonlinear system with unknown actuator dead-zone. In order to deal with this problem, a neural network controller is introduced to estimate the partial wind turbine system, and then another static network is constructed as a novel compensator to overcome the unknown dead-zone behavior of the actuator. The detailed design and analysis procedure of the proposed adaptive control scheme are presented and some statements are used to show the reasonability of the proposed method.