【摘要】 光伏发电的快速发展加剧了电力系统运行的相关性和不确定性.针对光伏发电中假设光伏出力服从Beta分布的局限性,提出三次样条函数通过有限矩对光伏出力进行重构,并应用改进的Nataf变换结合拉丁超立方采样技术得到计及相关性的光伏出力样本,通过分段线性化降低光伏出力波动性引起的误差,结合Cholesky分解求取输出量的半不变量,最后应用Cornish-Fisher级数求取输出变量的概率密度函数.该算法既计及光伏出力的相关性,同时也降低了传统半不变量法单点线性化引起的误差,与蒙特卡洛法(MCSM)相比,大大缩短了运算时间.对IEEE 34节点系统以及甘肃省会宁县某地区实际电网的测试验证了所提方法的准确性和实用性.更多还原

【Abstract】 With rapid development of photovoltaic power generation, the correlativity and uncertainty of power system operation are intensified. Aimed at the limitation of the assumption taken by photovoltaic power generation that its output should obey the Beta distribution, a cubic spline function is proposed in this article, the photovoltaic output is reconstructed by means of finite moment, the improved Nataf transformation combined with Latin hypercube sampling technique is used to fine the photovoltaic output sample on account of the correlation, the error caused by the fluctuation of photovoltaic output is reduced with piecewise linearization, the Cholesky decomposition is incorporated to find the cumulant of the output, and finally, the probability density function of the output variable is found by using Cornish-Fisher progression. The algorithm in this paper takes the correlation of the photovoltaic output into account as well as reduces the error caused by single-point linearization. Compared with Monte Carlo simulation method(MCSM), the operation time is greatly reduced. The accuracy and practicability of the proposed method are verified by means of testing the IEEE 34 bus system and wiring diagram of actual power grid in a region of Huining county, Gansu province.更多还原