光伏发电质量评价与惩罚函数研究

Research on quality evaluation and penalty function of solar power

文章基于光伏总出力、光伏出力波动和光伏出力预测误差3个评价指标,利用惩罚函数对光伏发电质量进行分析。首先,采用K均值聚类方法对上述指标的样本数据进行聚类;然后,采用F分布对聚类结果进行显著性检验,以提高聚类结果的有效性;最后,分别选取对数函数、幂函数、指数函数来建立惩罚函数,以量化各项评价指标对惩罚函数的影响程度,并综合采用层次分析法和熵权法来确定该惩罚函数的权系数。分析结果表明,文章所提出的光伏发电质量评价模型具有可行性以及较高的应用价值。

This paper studies the quality evaluation to the solar power, which is quantified by the penalty function considering the total output, the output fluctuation, and the forecast error. At first,the K-means clustering method is introduced to combine the sample data for each index. Then, the significance test using the F-distribution is applied to validate the effectiveness of the proposed model.Finally, the penalty terms based on the logarithmic function, the power function, or the exponential function are used to compare the influences of different indices on the penalty function. The Analytic Hierarchy Process(AHP) and the entropy method are used to determine the weight coefficients of the penalty function. The feasibility and application value of the proposed evaluation method are validated by the analytic results.