摘要
并行计算是实现大规模电力系统实时分析计算及控制的有效途径。将s级2s阶的辛龙格–库塔–奈斯通方法用于经典模型情况下的电力系统暂态稳定性计算中,利用矩阵分裂技巧以及矩阵求逆运算的松弛方法,推导出了一种新的暂态稳定性并行计算方法,该方法具有内在的时间并行特性和超线性收敛性。基于IEEE 145节点系统的仿真结果表明,该算法在保持相同或更高计算精度的前提下,具有与传统的时间并行严格牛顿计算方法相当的收敛性。
Parallel computation is an effective approach to real-time simulation and on-line control of large-scale power systems. Applying the S-stage 2S-order Symplectic Runge-Kutta-Nystrom method to the calculation of power system transient stability where the classic power system model is utilized and using the skill in Jacobian matrix splitting and the relaxation of matrix inversion, a new parallel computation method for power system transient stability is proposed. The proposed method possesses inherent parallel-in-time feature and super-linear convergence. Simulation results of IEEE 145-bus system show that under the precondition of keeping the same or higher computational accuracy, the proposed possesses the convergence equivalent to that of conventional parallel-in-time Newton approach using implicit trapezoidal rule.
出处
《电网技术》
EI
CSCD
北大核心
2011年第4期40-45,共6页
Power System Technology
基金
国家自然科学基金项目(50977052)~~
关键词
电力系统暂态稳定性
辛几何算法
并行算法
矩
阵分裂
松弛牛顿法
power system transient stability
symplectic algorithm
parallel algorithm
matrix splitting
relaxed Newton method
