10 Lidars and wind turbine control – Part 2

Modern utility-scale wind turbines are typically controlled using yaw, generator torque, and blade pitch actuation (Pao and Johnson, 2011). The yaw motor is used to align the rotor with the wind direction for the purpose of maximizing power capture. Generator torque is controlled during below-rated (region 2) operation to maximize power capture by maintaining the optimal tip speed ratio. In above-rated conditions (region 3), the blades are pitched to regulate rotor speed and power capture by limiting the amount of aerodynamic torque produced by the rotor. In addition to regulating rotor speed, region 3 control systems are often designed to reduce structural loads on the turbine caused by turbulence. A block diagram of a wind turbine’s feedback control system is shown in Fig. 131. In region 2, feedback measurements of generator speed are used to maintain the optimal tip speed ratio by adjusting generator torque. The generator torque command is traditionally set equal to the square of the generator speed multiplied by a constant whose value is chosen to maintain the optimal tip speed ratio (Pao and Johnson, 2011). If the wind speed is above the turbine’s rated wind speed, the region 3 controller employs blade pitch control to decrease the fraction of the available power in the wind extracted by the rotor. Feedback measurements of generator speed are input to a feedback blade pitch controller designed to both regulate rotor speed to the rated value and reduce structural loads caused by fluctuations in wind speed, which act as a disturbance to the wind turbine. With a single feedback measurement of generator speed, all of the blades are usually pitched together using a strategy called collective pitch (CP) control. When additional feedback measurements are available, such as the blade root bending moment at each blade, the blades can be pitched separately using individual pitch (IP) control (Bossanyi, 2005). Utilizing separate pitch commands for each blade allows for the reduction of loads due to variations in wind speed across the rotor disk, especially blade loads at the once-per-revolution (1P) frequency. A drawback to feedback control is that the wind disturbance must first act on the turbine before a corrective control action can be made based on a feedback measurement. To address this delay, feedforward control has been proposed, whereby a preview measurement of the