Research On Hybrid Energy Storage Control Strategy Based On Ultra-short-time Wind Power Prediction

In view of the randomness,volatility and uncertainty of large-scale wind power on the main network,a new control strategy based on the hybrid energy storage system of super capacitor and battery is proposed to stabilize the grid-connected power fluctuations.The paper uses the Fast Fourier Transform(FFT)to analyze the signal spectrum.A new method-Analytical Mode Decomposition(AMD)is used which decomposes the history wind power signal into high frequency and low frequency components based on the spectral characteristics and the compensation feature of the battery and super capacitor.The support vector machine method is used for ultra-short-term wind power prediction to forecast the low frequency components.The Gaussian radial basis kernel function is used in the ultra short term forecasting,in which the optimization of the penalty parameter and the kernel parameter is an important issue.Cuckoo algorithms is a new algorithm of heuristic,which can get the optimal solutions and don't need to look for all parameters,improving the learning capability of support vector machine.However,the search efficiency and accuracy will be reduced due to the random walk strategy and the fixed probability of the standard Cuckoo Search(CS)algorithm.This paper uses an Improved Cuckoo Search Algorithm(ICSA)method,the parameter-searched speed and accuracy is increased at the same time by using the adaptive dynamic control strategy instead of the fixed control strategy and the dynamic detection probability.After the ultra-short-term wind power forecasted for the low-frequency components,the grid-connected power fluctuation indicator of 1min and 30 min time-scale of the low-frequency is built.Whether the battery acts is determined by the fact that whether the power fluctuation indicator is out of control.Once any of the indicator is out of range,the battery acts.Then,the AMD method is adapted to adjust the cut-off frequency of the low frequency prediction signal until the requirements of grid-connected is met so as to decide the final compensate power of the battery.Last,the cut-off frequency of the original wind power is adjusted adaptively in the light of the SOC and compensation instructions of battery.The high frequency component is controlled by super capacity through fuzzy control.Based on the 99 MW wind farm in Inner Mongolia,the MATLAB simulation is analyzed based on the actual output power of wind farm data of 2014.The simulation result shows that this method can smooth power fluctuation,reduce the battery cycles,ensure the smooth ability of battery energy storage,avoid over charging and over discharging,prolong the life of battery.