风电出力的波动特性及预测方法研究

Fluctuation Characteristics and the Prediction Method of Wind Power

风能作为一种清洁能源,其有效利用对于全球能源互联网技术的发展具有重要意义。风电出力具有随机性、间歇性、波动性和不确定性的特点,这给电力系统的安全稳定运行及调度计划的合理制定等方面带来了挑战。基于大量实测数据,对风电出力的波动特性及其预测方法进行研究。首先,利用统计学的方法对风电出力在日内、日间、月度、季度等不同尺度下的平均值变化特点进行散点统计,并利用概率论对各时间尺度下的概率密度分布规律进行分析;其次,采用自回归模型与滑动平均模型相结合的时间序列法对风电出力进行短期预测。算例分析表明,风电出力具有不同时间尺度下的规律性,且文中所用预测方法误差较小,具有实用价值。

As a kind of clean energy, the wind power is of great significance to global energy internet. The wind power has many characteristics such as volatility, intermittency and uncertainty, which affect the safety operation and dispatching management of the power grid.Based on measured data, fluctuation characteristics and its prediction method of wind power are analyzed in this paper. Firstly, daily, monthly, and quarterly characteristics are analyzed by the statistics method, and their probability density figures are got by probability theory. Secondly, the wind power output is predicted by time series method which is combined by auto-regressive and moving average model. The example analysis shows the regularity of wind power output and small errors of prediction results, which indicate that this method is practical.