Secondary Voltage Control Strategy of Large Scale Renewable Energy Integrated Power Grid Based on RBF

摘要

Automatic voltage control (AVC) has been widely used in power grid dispatching system. With the high penetration of renewable generation, the reactive power uncertainty of source and load increases the voltage fluctuation, which brings new challenges to AVC. The traditional AVC strategy is based on the linear reactive voltage sensitivity method. When the system topology changes and the model update speed is not fast enough, the accuracy of AVC will be affected. This paper proposes a method for fitting the nonlinear relationship of reactive power and voltage based on historical data by Radial Basis Function (RBF) Neural Network, which can get more precise control even without precise parameters of model. We firstly obtain the voltage and reactive data of each node of the system with the inputs of the known generation and load data based on the simulation system. And then fit the nonlinear relationship between reactive power and voltage, which would replace the traditional secondary voltage control strategy. A 38 nodes sample system are applied to verify the control effects of the proposed method.