基于网络分块算法的静态安全快速计算方法

Fast static security calculation based on network partition algorithm

传统的静态安全分析需要对预想故障集中的预想故障进行多次网络重构和潮流计算,计算量和计算时间随电网规模的增大而快速增加,使得静态安全分析在实时分析和预警中不能满足全网分析的需求。提出一种基于网络分块算法的新型迭代方法,将全网合理划分为较为均匀的子网络,降低网络矩阵的维数,通过边界协调变量联系各子网络,有效提高计算速度。基于静态安全分析中故障潮流的特点提出适用于该算法的弱收敛性条件,该收敛条件使得每步迭代的计算量随迭代次数增加快速减少。仿真分析证明所提收敛条件容易达到,使得所提算法具有良好的收敛性。

Since multiple network reconfiguration and power-flow calculation should he carried out for each fault in the anticipated fault set during traditional static security analysis,the calculation amount and calculation time increase rapidly along with the expansion of power grid,which makes it impossible to satisfy the demand of grid-wide real-time analysis and warning. An iterative method based on network partitioning algorithm is proposed,which reasonably and quite evenly divides the whole grid into sub-networks to reduce the dimension of network matrix and coordinates the sub-networks via boundary variables to effectively enhance the calculation speed. A weak convergence condition is proposed according to the characteristics of the fault power-flow of static security analysis,which makes the calculation amount decrease rapidly along with the increase of iteration turns. Simulation verifies that,since the proposed convergence condition is easy to satisfy ,the proposed method has better convergence.