基于非参数核密度估计法的光伏出力随机分布模型

Random Distribution model of Photovoltaic Output Based on Non-parametric Kernel Density Estimation Method

光伏出力的随机性是影响光伏发电稳定性的重要因素。通过对国内某一光伏电站历史运行数据进行统计计算和分析,揭示气象条件差异对光伏出力水平的影响规律,并提出将数据按气象条件分类,用非参数核密度估计法建立晴天、多云、阴天、雨雪4种气象条件下光伏出力的随机分布模型。该方法对数据分布不附加任何假定,只从数据本身特征出发研究其分布特性,适合分析受气象影响显著的光伏出力分布特性。结果表明,模型所示分布特性与相应气象条件下光伏出力实际概率分布具有很好的一致性。

The random characteristics of photovoltaic(PV) output is one of the significant factors affecting the stability of PV power generation.The influence of weather difference on the level of PV output was studied by statistical calculations and analysis of the historical operating data of a PV power plant.Meanwhile,the random distribution models of PV output were established under clear,cloudy,overcast and rainy or snowy weather conditions by non-parametric kernel density estimation method.The non-parametric kernel density estimation method can analyze the distribution features of the data without any additional assumptions,which is suitable to analyze the random distribution characteristics of PV output that is influenced significantly by weather conditions.The analysis indicates that these models are well consistent with the actual probability distribution of single day PV output under corresponding weather conditions.