A new control strategy of energy storage systems (ESSs) based on wind power prediction deviation and battery state of charge (SOC) feedback is proposed. The deviation of wind power variation is calculated through prediction results to get the charge-discharge power of ESSs needed by totally compensating fluctuation. The charge-discharge power orders are then obtained by introducing the compensation factor in the joint solution. Moreover

a dynamic optimizing model of the compensation factor is developed

including the standard compensation factor optimizing model based on output power fluctuation and battery capacity changing value and the compensation factor fast-modification model based on battery SOC and charge-discharge status under short-time scale. The optimal solution and SOC value used in the proposed algorithm have a high adaptation level

which can generalize ESSs of different capacities in the application of wind power smoothing and give consideration to the lifetime of battery and smoothness of wind power. Finally

a case study is made that describes simulations with the historical power data on wind farm and ESSs configuration taken into account. The results have proved that the control strategy is able to develop the capacity of ESSs to the full and reduce the fluctuation of wind farm power on the premise that the energy capacity of battery is maintained stable.