基于预测误差分布特性的风电场短期功率预测不确定性研究

UNCERTAINTY ANALYSIS OF SHORT TERM WIND POWER FORECASTING BASED ON ERROR CHARACTERISTICS STATISTICS

通过研究预测误差的分布特性,提出一种预测评价方法。将预测误差按风电场功率值大小划分区间,在已建预测模型基础上应用蒙特卡罗随机模拟原理,同时考虑预测误差对预测结果的影响,进行风电场功率短期预测及不确定性分析。将其应用于中国北方某风电场,经过数据方法的有效性检验,证明了该方法的实用性,为风电场功率预测系统的可靠性分析提供支持,提供了一种新的思路。

Wind power output forecasting plays crucial part in wind power integration and dispatch. Evaluation and analysis of forecasting results are important to reduce the errors. By means of studying statistic characteristics of forecasting errors, a method was proposed to test the suitability of forecasting model. Forecasting errors were divid- ed into several intervals according to wind power output, and the effects of errors on results was considered, Monte Carlo method was utilized to analyze uncertainties of built forecasting model. After applying this method into a wind farm in North China, its practicability is verified, which supports the reliability of wind power forecasting system.