基于Nataf变换含相关性的扩展准蒙特卡洛随机潮流方法

发布时间：2018-01-22 23:32

本文关键词： 随机潮流 Nataf变换扩展准蒙特卡洛方法奇异值分解相关系数矩阵　出处：《电工技术学报》2017年02期 　论文类型：期刊论文

【摘要】：随机潮流分析中,准蒙特卡洛方法在同样规模下不仅计算效率高于基于拉丁超立方的方法,且具有更好的扩展性质。因此提出一种基于Nataf变换的扩展准蒙特卡洛方法(NEQMC)并应用于概率潮流计算中。该方法利用Nataf变换重构输入变量的概率分布,而扩展技术可在随机潮流未收敛时保留已知的潮流计算结果,基于奇异值分解的相关系数控制技术不仅可保持扩展前后样本的相关性,且使得该方法也适用于相关系数矩阵非正定的情况。与基于Nataf变换的简单随机抽样、扩展拉丁超立方方法在IEEE 30和IEEE 118节点系统的比较分析证明了所提方法的高效性和准确性,仿真结果表明:相比于拉丁超立方和简单随机抽样,NEQMC的输出变量准确度更高,特别是标准差的准确度得到大幅改进,获得相同准确度的计算复杂度大大减小。
[Abstract]:In stochastic power flow analysis, the quasi-Monte Carlo method is not only more efficient than the Latin hypercube method in the same scale. Therefore, an extended quasi-Monte Carlo method based on Nataf transform is proposed. The method uses Nataf transform to reconstruct the probability distribution of input variables. The extended method can keep the known power flow calculation results when the stochastic power flow is not convergent, and the correlation coefficient control technique based on singular value decomposition can not only keep the correlation of the samples before and after the expansion. The method is also suitable for the case of non-positive definite correlation coefficient matrix and simple random sampling based on Nataf transform. The efficiency and accuracy of the proposed method are proved by comparing and analyzing the extended Latin hypercube method in IEEE 30 and IEEE 118 bus systems. The simulation results show that the accuracy of the output variables of NEQMC is higher than that of Latin hypercube and simple random sampling, especially the accuracy of standard deviation is greatly improved. The computational complexity of obtaining the same accuracy is greatly reduced.
【作者单位】：上海交通大学电子信息与电气工程学院;中国电力科学研究院;
【基金】：国家自然科学基金重点项目(51337005) 国家重点基础研究发展计划(973计划)(2014CB23903)资助
【分类号】：TM744
【正文快照】： 0引言电力系统运行受到各种随机因素的影响[1],为有效评估其潜在运行风险,B.Borkowska于1974年提出随机潮流(Probabilistic Load Flow,PLF)[2]。经过数十年发展,目前PLF在静态分析中应用广泛[2-12]。蒙特卡洛方法准确度高且不受系统非线性的影响,是PLF的重要方法[10-12]。但基

【相似文献】