Research Of Power Smoothing Technology For Wind Farm On Energy Storage

In recent decades, the energy issue gradually attracts the world’s attention. Characteristics of non-renewable and limitation of petroleum, coal and other fossil fuels make countries to consider developing clean, renewable energy vigorously. The development of renewable clean energy is raised to a political height in China. Among all the clean energy, wind energy is abundant. The wind power technology is relatively mature. Compared to traditional energy, wind energy has no by-product contaminants and it is renewable. Along with the development of wind power technology, the development of wind power is strong and the installed capacity continues to increase in China. However, there are some inherent characteristics, such as random, intermittent, instability in wind power which will be a threat to the safety of large power grids, So the research of smoothing wind power fluctuations and storage capacity configuration is particularly important.In this paper, the main work includes the following sections:(1)A high precision wind speed prediction method is proposed. On the basis of analysis of the basic principles of the least squares support vector machine(LSSVM) regression model and quantum particle swarm optimization(QPSO) algorithm, the short-term values are predicted based on historical data. In the process of LSSVM parameters optimization QPSO algorithm is introduced, QPSO-LSSVM model is introduced to the wind speed forecasting first time. Finally, comparising results of several other prediction algorithm is verified in MATLAB.(2)A control strategy for smoothing wind power fluctuations is proposed.This strategy calculates wind power according to wind turbine curve on the basis of wind prediction. Then the prediction result power values are treated as input and the moving average method is used to smooth power fluctuations in advance. Window size alters in the moving average algorithm The change of window size considers the volatility of wind power and SOC of energy storage systems. Finally, the control strategy is verified.(3)A storage capacity configuration method considering the overall economic efficiency is proposed. This method analyzes the impact of storage capacity configuration on power smoothing. The battery charge and discharge model are established. The constraint function, objective function and evaluation indicator in the case of considering operating costs are proposed. Then adoptive QPSO algorithm is introduced to solve the objective function of battery pack. Finally, the smoothing effect of the optimal configuration at different time windows is studied without considering operating costs and full capacity configuration.