Load control optimization method for offshore wind turbine based on LTR

Abstract

A simple superposition method of two separate pitch control loops is adopted in the traditional tower load control and output power control, ignoring the coupling of rotor rotation and tower movement under aerodynamic force. Besides, tower load control performance is dependent on the accuracy of the tower motion sensor, resulting in a decrease in system reliability. Therefore, a pitch controller without sensor is proposed to achieve the coordinated control of tower load and output power. First, a low-order wind turbine state–space model is derived which is suitable for the design of pitch controller and compared with the linearized model from GH Bladed to verify the rationality of the deduced model. Then, based on the theory of multi-input and multi-output (MIMO) control, a linear quadratic Gaussian (LQG) pitch controller is designed integrated with disturbance accommodation controller (DAC). The Loop transfer recovery (LTR) method is used to restore dynamic characteristics of closed-loop to improve the performance of the pitch controller. The feasibility of the LTR method is proved by detailed theoretical analysis. Finally, compared with traditional controllers, it is verified that the proposed LTR pitch controller can further reduce tower load while stabilizing output power in different wind conditions.

Keywords

• Pitch control
• Tower load
• Linear quadratic Gauss control
• Kalman filter
• Loop transfer recovery