基于稀疏矩阵变换的电网故障电流快速算法  

Fast Algorithm for Power Grid Faulted Current Based on Sparse Matrix Transform

作　　者：陈志光 范幸[2] 李一泉 朱峥 邱建 黄明辉 

机构地区：[1]广东电网有限责任公司电力调度控制中心,广东广州510000 [2]杭州智光一创科技有限公司,浙江杭州310058

出　　处：《广东电力》2015年第9期50-55,共6页Guangdong Electric Power

基　　金：国家自然科学基金资助项目(61233008);广东电网有限责任公司科技项目(K-GD2014-099)

摘　　要：结合稀疏矩阵低阶更新算法和低秩分解理论,提出一种电网故障电流快速算法。首先,基于对称分量法将故障引起的网络结构变化用经过相序变换的修正导纳矩阵表示;然后,根据非负矩阵分解原理将其变换成2个低秩矩阵乘积的形式;最后,利用已形成的因子表,采用低阶更新算法实现故障后网络电流的快速计算。新算法充分利用了电力网络节点导纳矩阵的稀疏特征,对其进行低阶分解处理,大大减少了等值阻抗的计算量,提高了短路电流的计算速度。分别在IEEE-14、IEEE-30、IEEE-118和IEEE-1047节点系统对新算法进行测试,并与传统的修正导纳矩阵法进行对比,结果表明新方法的计算精度能满足工程要求,在大规模系统中计算速度的优势明显。Combining sparse matrix low rank update algorithm and low rank decomposition theory,a kind of fast algorithm for power grid faulted current was proposed.Firstly,change of network structure caused by fault was shown by modified admittance matrix after phase-sequence transformation.Then,according to non-negative matrix decomposition principle,the modified admittance matrix was transformed to form of product of two low rank matrix.Finally,the formed factor table and low rank update algorithm was used to realized fast calculation on network current after the fault.The new algorithm fully used sparse characteristic of admittance matrix of electric power network nodes to deal with low rank decomposition which greatly reduced calculated amount for equivalent impedence and improved calculation speed for short circuit current.IEEE-14 node system,IEEE-30 node system,IEEE-118 node system and IEEE-1047 node system was respectively used for testing the new algorithm and compared results of the new algorithm with those of traditional modified admittance matrix methods.Results indicates that calculation precision of the new method could satisfy engineering requirements which is of obvious advantage in calculation speed for large scale system.

关 键 词：稀疏矩阵 低阶更新 低秩分解 修改导纳 故障电流 

分 类 号：TM711.2[电气工程—电力系统及自动化]