分布函数差异化导向的风电功率预测误差气象条件概率建模方法

Probabilistic modeling method of weather condition for wind power forecasting error based on differentiation orientation of distribution function

从统计学角度分析条件概率分布的差异性对于随机决策问题的重要性,并提出一种分布函数差异化导向的风电功率预测误差气象条件概率建模方法。以预测误差概率分布总体差异程度最大化为目标,利用改进的K-means算法进行气象数据聚类,并基于聚类结果对风电功率预测误差数据进行分箱;采用通用分布拟合不同气象模式的误差概率密度,得到解析化的风电功率预测误差气象条件概率分布。利用支持向量机实现气象模式识别,从而基于数值天气预报为调度计算提供相应气象模式下的误差条件概率模型。以中国华北地区某风电场历史数据为例,验证了所提方法的有效性,且相较于不考虑气象的简单统计频次模型,所提模型可使系统的风电消纳水平得到有效提升。

The importance of the difference of conditional probability distribution to stochastic decisionmaking problem from the perspective of statistics is analyzed,and a probabilistic modeling method of weather condition for wind power forecasting error is proposed based on differentiation orientation of distribution function.A modified K-means algorithm is used for clustering the weather data,which takes the maximum total difference of probability distribution of forecasting error as its objective,and the forecasting error data of wind power is binned based on the clustering results.The versatile distribution is adopted to fit the probability density of the errors for different weather modes,and the analytical probability distribution of weather condition for wind power forecasting error is obtained.The support vector machine is adopted to achieve the recognition of weather modes,further the conditional probability models of the errors under corresponding weather modes are provided for the scheduling calculation based on the numerical weather prediction.The historical data of a wind farm in North China is taken as an example,the effectiveness of the proposed method is verified,and the proposed model can effectively improve wind power consumption level of the system compared with the simple statistical frequency model without considering weather.