贾小凡, 叶华, 刘玉田, 刘牧阳. 基于PIGD-IRK的大规模时滞电力系统特征值高效计算方法[J]. 电力系统自动化, 2023, 47(12): 95-102.
引用本文: 贾小凡, 叶华, 刘玉田, 刘牧阳. 基于PIGD-IRK的大规模时滞电力系统特征值高效计算方法[J]. 电力系统自动化, 2023, 47(12): 95-102.
JIA Xiaofan, YE Hua, LIU Yutian, LIU Muyang. High-efficiency Eigenvalue Calculation Method for Large-scale Time-delay Power System Based on Partial Infinitesimal Generator Discretization with Implicit Runge-Kutta[J]. Automation of Electric Power Systems, 2023, 47(12): 95-102.
Citation: JIA Xiaofan, YE Hua, LIU Yutian, LIU Muyang. High-efficiency Eigenvalue Calculation Method for Large-scale Time-delay Power System Based on Partial Infinitesimal Generator Discretization with Implicit Runge-Kutta[J]. Automation of Electric Power Systems, 2023, 47(12): 95-102.

基于PIGD-IRK的大规模时滞电力系统特征值高效计算方法

High-efficiency Eigenvalue Calculation Method for Large-scale Time-delay Power System Based on Partial Infinitesimal Generator Discretization with Implicit Runge-Kutta

  • 摘要: 电力系统控制回路中固有的时滞会影响控制器的性能,甚至危害系统的稳定运行。为此,提出了一种基于无穷小生成元隐式龙格-库塔部分离散化(PIGD-IRK)的特征值计算方法,用以分析时滞对大规模电力系统小干扰稳定性的影响。首先,分析了制约无穷小生成元隐式龙格-库塔离散化(IGD-IRK)方法计算效率的瓶颈;然后,应用部分谱离散化思想大幅度降低其构造的无穷小生成元离散化矩阵的维数;接着,利用离散化矩阵的结构特点将其从阶梯型分块矩阵相似变换为常数矩阵的逆和分块上三角矩阵的乘积,从而可在特征值计算过程中充分利用离散化矩阵和系统增广状态矩阵的稀疏特性。最后,通过四机两区域系统和2个实际电网的计算结果验证了所提方法的准确性和高效性。

     

    Abstract: The inherent time delays in the control loops of power systems will affect the performances of controllers and even endanger the stable operation of the system. Therefore, an eigenvalue calculation method based on the partial infinitesimal generator discretization with implicit Runge-Kutta(PIGD-IRK)is proposed for analyzing the effect of time delays on small disturbance stability of large-scale power systems. Firstly, the bottlenecks that restrict the computational efficiency of infinitesimal generator discretization with implicit Runge-Kutta(IGD-IRK)are analyzed. Secondly, by applying the partial spectral discretization idea, the dimension of the infinitesimal generator discretization matrix constructed by IGD-IRK is greatly reduced. Thirdly, the echelon block discretization matrix is further transformed into the product of the inverse of a constant matrix and an upper triangular block matrix by taking advantage of its structural characteristics, thus the sparsity of the discretization matrix and system augmented state matrix can be fully used in the process of eigenvalue calculation. Finally, the accuracy and high efficiency of the proposed method are verified on a two-area four-machine system and two actual power systems.

     

/

返回文章
返回