Probabilistic load flow for photovoltaic distributed generation using the Cornish–Fisher expansion
Probabilistic load flow for photovoltaic distributed generation using the Cornish–Fisher expansion
复制标题
DOI：
10.1016/j.epsr.2012.03.009
复制
发表时间：
2012-08
影响因子：
3.9
通讯作者：
F. Ruiz-Rodriguez;J. Hernandez;F. Jurado
中科院分区：
文献类型：
--
作者：
F. Ruiz-Rodriguez;J. Hernandez;F. Jurado
本文表明，为了解决辐射状配电网的概率潮流，有必要采用有效的技术，考虑到他们的技术约束。在这些限制中，电压调节是光伏分布式发电中要解决的主要问题之一。概率潮流可以用解析法和蒙特卡罗法求解。我们的研究应用了一种分析方法，结合累积量方法和Cornish-Fisher展开来解决这个问题。用蒙特卡罗方法对解析方法的计算结果进行了比较。为了评估光伏分布式发电的性能，本文描述了一个概率模型，考虑到随机性质的太阳辐照度。因此，负荷和光伏分布式发电建模为独立/相关的随机变量。所得结果表明，该技术提出了一个更好的性能比蒙特卡罗方法。这种技术提供了令人满意的解决方案，迭代次数较少。因此，收敛速度很快，计算成本低于蒙特卡洛方法所需的。此外，研究结果还揭示了当输入随机变量为非高斯分布时，Cornish-Fisher展开比Gram-Charlier展开具有更好的性能。
This paper shows that in order to solve a probabilistic load flow in radial distribution networks, it is necessary to apply effective techniques that take into account their technical constraints. Among these constraints, voltage regulation is one of the principal problems to be addressed in photovoltaic distributed generation. Probabilistic load flows can be solved by analytical techniques as well as the Monte Carlo method. Our research study applied an analytical method that combined the cumulant method with the Cornish–Fisher expansion to solve this problem. The Monte Carlo method is used to compare the results of analytical method proposed. To evaluate the performance of photovoltaic distributed generation, this paper describes a probabilistic model that takes into account the random nature of solar irradiance. Therefore, load and photovoltaic distributed generation are modelled as independent/dependent random variables. The results obtained show that the technique proposed gave a better performance than the Monte Carlo method. This technique provided satisfactory solutions with a smaller number of iterations. Therefore, convergence was rapidly attained and computational cost was lower than that required for the Monte Carlo method. Besides, the results revealed how the Cornish–Fisher expansion had a better performance than the Gram–Charlier expansion, when input random variables were non-Gaussian.