基于改进拉丁超立方抽样的概率潮流计算

Evaluation of Probabilistic Load Flow Based on Improved Latin Hypercube Sampling

针对基于拉丁超立方抽样的蒙特卡洛模拟法应用于概率潮流计算时相关性控制过程存在的问题,提出了一种新的可以精确处理随机变量相关性的方法。为了使采样值更为精确地反映随机变量的数字特征,在正态分布变量的样本生成过程中引入区间均值采样方法。根据离散分布的特点,在离散分布随机变量样本生成过程中引入了离散拉丁超立方抽样方法。本文提出的方法不仅可以提高计算精度,同时可以全面地给出输出变量的数字特征以及分布。IEEE 30节点仿真结果验证了所提出方法的有效性。

In the light of problems in handling correlation in probabilistic load flow（PLF） calculation with Latin hypereube sampling based on Monte Carlo simulation, a new method that can accurately handle dependencies among variables is proposed. Improved sampling of interval mean value that can reflect digital features of random variables more precisely is introduced to the process of sample generation of normal distributed variables. In accordance with the characteristics of discrete variables, discrete Latin hypercube sampling is employed to produce the samples of discrete variables. The proposed method can not only improve the accuracy, but also present the digital features and the probability distribution of the output variables comprehensively. Case studies carried on the IEEE 30-bus system verify the effectiveness of the proposed method.