引用本文: | 彭寒梅,曹一家,黄小庆,罗 聪.基于组合抽样的含分布式电源随机潮流计算[J].电力自动化设备,2014,34(5): |
| PENG Hanmei,CAO Yijia,HUANG Xiaoqing,LUO Cong.Probabilistic load flow calculation based on combination sampling for power system containing distributed generations[J].Electric Power Automation Equipment,2014,34(5): |
|
摘要: |
针对含分布式电源系统随机潮流计算中随机变量类型多的特点,提出基于组合抽样的蒙特卡罗模拟法,其根据随机因素的不同特点采用不同的抽样方法,并计及线路随机故障的影响。将随机因素作为随机变量建立概率分布模型;对连续概率分布的负荷及分布式电源出力采用拉丁超立方抽样,对离散概率分布的电源和线路随机故障采用结合拉丁超立方抽样的改进重要抽样方法,以减小方差,提高模拟计算效率;对系统随机变量进行抽样建模。通过对接有分布式电源的IEEE 14节点系统和IEEE 57节点系统的算例分析,表明了所提方法是有效的,并具有良好的收敛性。 |
关键词: 随机潮流 分布式电源 组合抽样 拉丁超立方抽样 重要抽样 模型 |
DOI: |
分类号: |
基金项目:国家高技术研究发展计划(863计划)资助项目(2012AA050217) |
|
Probabilistic load flow calculation based on combination sampling for power system containing distributed generations |
PENG Hanmei, CAO Yijia, HUANG Xiaoqing, LUO Cong
|
College of Electrical and Information Engineering,Hunan University,Changsha 410082,China
|
Abstract: |
Since there are multiple random variables in the probabilistic power flow calculation of power system containing distributed generations,a Monte Carlo simulation method based on combination sampling is proposed,which applies different sampling methods according to the different characteristics of random factors and considers the influence of random line fault. A probability distribution model is built with the random factors as random variables. In order to reduce the variance and improve the calculation efficiency,Latin hypercube sampling is adopted for loads and distributed generation outputs,which have continuous probability distribution,while the improved importance sampling method combined with Latin hypercube sampling is adopted for random power source and lines faults,which have discrete probability distribution. The sampling models are established for system random variables. IEEE 14-bus system and IEEE 57-bus system containing distributed generations are analyzed and results show that the proposed method is effective and has good convergency. |
Key words: probabilistic load flow distributed power generation combination sampling Latin hypercube sampling importance sampling models |