Simulation of Wind Power System Involving Flywheel Energy Storage Unit Based on Wind Speed Forecasting by RBF Neural Network

Yi Feng,Heyun Lin,Jianhu Yan,Xiaoquan Lu,Hui Yang

DOI: https://doi.org/10.1109/icrera.2013.6749855

2013-01-01

Abstract:A wind power system (WPS) involving a flywheel energy storage (FES) unit is proposed to reduce the fluctuation of output power due to the intermittent nature of wind power. The control of the FES unit is based on a wind speed forecasting technique by the RBF neural network. A new control strategy adopting different reference DC bus voltages under different forecasting wind speeds is presented to improve the output power regulation capability of the FES unit. The wind speeds in the next time array can be accurately forecasted by the radical basis function (RBF) neural network. The mathematical model of a stand-alone wind power generation system including a FES unit is established and simulated by Matlab/Simulink. The simulation results verify the correctness of the system model and the effectiveness of the control method.