Sag detection algorithm for dynamic voltage restorer used in wind farms under unbalanced and distorted grid voltage conditions

Dynamic voltage restorer (DVR) is a cost effective solution to solve the grid voltage sag problem. Recently DVR is used to fulfill LVRT capability for DFIG wind power systems. The terminal voltage of wind farms usually has large fluctuation and is distorted by voltage harmonics when connected to weak grid. This demands a high requirement for sag detection algorithm, which should detect the voltage sag fast and avoid being falsely triggered by the disturbance. Through analyzing the grid fault, in this paper it is concluded that both the positive and negative sequence component should be considered in the sag detection criterion. Then this paper proposes a more generalized sag detection criterion, which combines the positive and negative sequence component in a linear relationship. The recursive discrete Fourier transform (rDFT) method with 10ms moving window is used to extract the sequence component magnitude, which provides good robustness to grid voltage harmonics and doesn´t deteriorate the response time. Simulation is implemented to verify the validity of the proposed sag detection algorithm. The simulation result shows that the proposed method has a good performance under unbalanced and distorted grid voltage conditions. The effect of the phase jump is also considered. Compared with other sag detection method, the proposed sag detection algorithm can response much fast and eliminate the influence of voltage harmonics.