含风电的电力系统概率潮流计算

Probabilistic Load Flow Calculation for Power Grid Containing Wind Power

概率潮流求解中蒙特卡罗法只有在大规模采样的条件下进行多次模拟,才能提高精准度,其导致计算量大,耗费时间,难以处理风电中变量相关性的概率潮流。采用基于拉丁超立方采样的蒙特卡罗法对含有风电场的电力系统概率潮流问题进行分析。基于拉丁超立方采样的蒙特卡罗法,主要分为采样和排序。采样是为了确保样本空间能够被完整的采样,排序是为了降低随机变量之间的相关性。该方法将Gram-Schmidt和Cholesky2个排序方法结合,很好地降低随机变量之间的相关性。通过IEEE-39节点仿真,结论显示该方法能够较好地处理风电中的风速相关性,降低采样规模,提高精准度,是一种非常有效的处理含有风电场的概率潮流问题的方法。

The Monte Carlo method in the probabilistic load flow solution improves the accuracy only under the condition of large quantities of samplings and many times of simulations,leading to large amounts of calculation and time consuming,therefore it is difficult to deal with variable correlation in the wind power probabilistic load flow solving. In this paper,we use the Monte Carlo method based on Latin hypercube sampling to analyze the probabilistic load flow of the power system containing wind power. The Latin hyper cube sampling contains the two processes of sampling and sorting. The sampling is to ensure that the sample space can be completely sampled and the sorting is to reduce the correlation between random variables. This method integrates two sorting methods： Gram-Schmidt and Cholesky,so as to be able better reduce the correlation of random variables.The results of the IEEE-39 node simulation shows that this method can preferably deal with the correlation of the wind speed,reduce the sampling size and improve accuracy and is a very effective way to deal with the method of probability flow problem containing the wind farm.