基于输入变量秩相关系数的概率潮流计算方法

随着大量新能源接入,电力系统运行必须考虑其随机性带来的影响。概率潮流是有效工具之一。针对考虑输入变量相关性的概率潮流计算,文中采用Spearman秩相关系数表示输入随机变量间的相关性,分析了拉丁超立方抽样方法与秩相关系数的内在关联,提出结合遗传算法的改进拉丁超立方抽样方法进行概率潮流计算。算例结果表明,所提出的方法能较好地刻画风速间的相关性,不受输入随机变量边缘分布的影响,并且能处理秩相关系数矩阵正定和非正定的情况。

Probabilistic Load Flow Calculation Based on Rank Correlation Coefficient of Input Random Variables

With numerous new energy resources integrated into the power system, the influences brought about by random variables have to be properly considered in the operation of power systems. Probabilistic load flow is one of the effective tools. In this paper, the method for probabilistic load flow considering the dependence among variables is studied. The Spearman rank correlation coefficient is used to model the dependence among variables, and the inherent relation between Latin hypercube sampling and rank correlation coefficient is analyzed. Latin hypercube sampling combined with genetic algorithm is proposed to solve probabilistic load flow. Simulation results show that the method has a better performance than others in describing the dependence between wind speeds, and is not influenced by different marginal distributions. Moreover, it can handle positive and non-positive rank correlation coefficient matrices.