Nonlinear Model Predictive Control of Wind Farm for System Frequency Support

Abstract

The injection of a significant amount of wind power tends to increase the rate of change of the grid frequency. Therefore, it is a trend for wind farm to participate in the grid frequency regulation. However, during the frequency control process, individual wind generators in a wind farm may prone to instability due to possible over-deceleration. To address this issue, this paper presents a new nonlinear model predictive control (NMPC) scheme for the wind farm frequency response. By incorporating the nonlinear dynamics of each individual wind generator into the NMPC design, it achieves both objectives of dynamically optimal frequency response and wind generator stability. This scheme has a three-layer structure. Based on the linear model predictive control and moving horizon estimation, a top-layer controller computes the overall wind farm power reference to support the frequency control. This overall power reference is fed to the middle-layer NMPC, and is further distributed among multiple wind generators. The dispatched power references are then sent to the bottom-layer wind generators local controllers for execution. Simulation results verify the effectiveness of the proposed scheme.