基于RBF神经网络的光伏并网系统自适应等效建模方法OA北大核心CSTPCD

There is a problem that the PV grid-connected system model in the generalized load modeling is difficult to adapt to the dynamic response of different inverter control and frequency disturbance.Thus this paper proposes an adaptive equivalent modeling method for a PV grid-connected system based on a radial basis function(RBF)neural network.First,the detection criteria of response waveforms of different control strategies of photovoltaic grid-connected inverters are established.Second,an RBF neural network model is constructed with voltage and frequency disturbances as input and active and reactive power as output.Finally,a photovoltaic grid-connected system model is built in Matlab/Simulink and connected to the IEEE14 node distribution network for simulation verification.The results indicate that the constructed adaptive equivalent model can effectively identify the types of voltage and frequency control,active and reactive power control,and droop control strategies,and can accurately reflect the dynamic response characteristics of the photovoltaic grid-connected system's active and reactive power under different voltage and frequency disturbances.