A Control Strategy for Battery Energy Storage Smoothing Short-Term Wind Power Fluctuation Based on ANN

With the Grid-connected scale of wind generators increasing, the random fluctuation of wind power has been one of the critical issues for the consumption of wind power. From the perspective of how to use the battery energy storage system (BESS) to smooth the wind power fluctuation, the control scheme for two BESSs using the alternative charging/discharging strategy to smooth the wind power fluctuation is proposed at first. Then, the design method for determining the capacity of the two BESSs based on the statistics of wind power fluctuation is studied. Due to the adaptive neural network (ANN) has the excellent ability in setting up the non-linear mapping between the input and output of the controlled object whose model is difficult for building, the ANN is therefore used to generate the reference command for BES Ss smoothing the short-term and strongly random wind power fluctuation. At the same time, the improved method for adaptively adjusting the weight of neurons based on modifying deviation latch is proposed to improve the accuracy of ANN on predicting the short term wind power fluctuation. Finally, both the feasibility of both the proposed charging and discharging strategy in saving the battery cycle life and the effectiveness of the proposed ANN control strategy in smoothing the short-term wind power fluctuation is validated by simulating in the improved 11-bus wind power system.