Dual-Stage MPC-Based AGC for Wind Farm Considering Aerodynamic Interactions

Abstract

The wind farms are encouraged to provide Automatic Generation Control (AGC) services for the power grid. However, the complex aerodynamic interactions between turbines complicate the control of wind farms for AGC service. To address this issue, this paper proposes an explicit wind speed prediction model of each wind turbine based on the control actions of thrust coefficient (mainly realized by adjusting pitch angle) and yaw angle. Its accuracy is validated by numerical experiment based on Navier-Stokes equations. Since the adjustments of the thrust coefficients and yaw angles involve significantly different time scales, a dual-stage model predictive control (MPC) is proposed to coordinate them. It uses the proposed explicit wind speed prediction model as a surrogate for the wind speed of each turbine. In the first stage, it optimizes yaw angle reference based on whether or not the available wind farm power is sufficient at the predicted moments. In the second stage, it controls both thrust coefficient and yaw angle of each turbine, in order to track the AGC power signal and align the yaw angles with those determined in the first stage. Case studies demonstrate the effectiveness of the dual-stage MPC for AGC power tracking of wind farm.