考虑光伏电站时空相关性的光伏出力序列生成方法

发布时间：2018-05-22 13:45

本文选题：光伏发电 + 随机时间序列生成　；参考：《中国电机工程学报》2017年07期

【摘要】：为了更好地研究光伏并网对电力系统的影响,该文提出一种考虑时空相关性的光伏出力序列生成算法。首先针对国内外多个光伏电站的实测出力曲线进行分析,结合太阳辐射规律性变化趋势、每日总体大气衰减情况、局部云层扰动等影响光伏出力的3个主要因素,将光伏出力分解为理想出力归一化曲线、幅值参数和随机分量3种成分。在此基础之上,提出利用典型日数据的理想出力提取方法和基于向量自回归模型(vector auto-regressive,VAR)的幅值参数序列生成方法。前者能够精确描述光伏出力的规律性变化趋势,后者不仅可有效保持单光伏电站的出力均值、方差、概率分布等重要统计特性,还能较好地表征多个光伏电站之间的空间相关性。随机分量经验证满足t location-scale分布,可通过蒙特卡罗抽样生成。最后,利用甘肃6个光伏电站及德国5个区域级光伏出力数据对所提方法的有效性进行仿真验证。
[Abstract]:In order to better study the influence of photovoltaic (PV) grid connection on power system, this paper proposes an algorithm for generating photovoltaic force series considering the temporal and spatial correlation. Firstly, the measured output curves of many photovoltaic power stations at home and abroad are analyzed, and three main factors affecting photovoltaic output are analyzed, such as the trend of solar radiation regularity, the daily total atmospheric attenuation, the local cloud disturbance, and so on. The photovoltaic force is decomposed into three components: ideal force normalization curve, amplitude parameter and random component. On the basis of this, an ideal extraction method based on typical daily data and a method to generate amplitude parameter sequences based on vector autoregressive model (vector auto-regressive VAR) are proposed. The former can accurately describe the regular trend of photovoltaic output, while the latter can not only effectively maintain the mean, variance, probability distribution and other important statistical characteristics of a single photovoltaic power plant. The spatial correlation between several photovoltaic power stations can also be well characterized. It is proved that the random component satisfies t location-scale distribution and can be generated by Monte Carlo sampling. Finally, the validity of the proposed method is verified by using six photovoltaic power stations in Gansu and five regional PV output data in Germany.
【作者单位】：强电磁工程与新技术国家重点实验室(华中科技大学);ABB(中国)有限公司 ABB中国研究院;
【基金】：国家自然科学基金项目(51577075) 国家重点研发计划智能电网技术与装备专项项目(2016YFB0900400,2016YFB0900403)~~
【分类号】：TM615

【参考文献】