An Early Warning Method Based on Improved Affine Arithmetic for Wind Power Ramp Events in Power System

Abstract

With more and more wind power integrated into power system, increasing attention is paid to wind power ramp events (WPREs) due to the significant impact on the power balance of power system. Considering the impact of large volatility of WPREs on power control, the permissible intervals of ramping amplitude are analysed to determine the boundaries between different warning stages, which are corresponding to different power control measures to keep the power balance with Area Control Error (ACE) deviation maintained within the allowable range. Moreover, interval analysis is introduced to deal with the strong uncertainty, and an improved affine arithmetic is proposed to deal with the conservation brought by interval analysis by considering the temporal correlation of WPREs prediction error. A multi-stage early warning method for WPREs is established in further, by which the warning information, including boundaries and probability distributions, can be provided reflecting the degree of potential harm inflicted by WPREs. Simulation results demonstrate that, the provided waring information have clear physical meanings to provide guidance for the operators, and the proposed warning method has simple and rapid characteristics, which reflects the verify the effectiveness of the proposed method.