摘要
对最优潮流牛顿算法实际应用中的两个主要困难：数值不稳定性和如何快速正确地预估起作用的不等式约束集分别作了详尽的分析研究，并提出了相应的处理对策。包括：一种改进的软惩罚方案；一种根据电网拓扑结构的快速预估起作用的电压不等式约束集的方案；试验迭代的有限次终止方案等等。并对多个ＩＥＥＥ 试验电力系统和我国２ 个实际电力系统作了全面的数值计算。结果表明，上述改进措施和方案，显著地提高了牛顿ＯＰＦ算法的数值稳定性、收敛性和计算速度，具有实用意义。
In this paper, some strategies dealing with two main difficulties of Optimal Power Flow Newton approach, i.e. the numerical instability and the identification of binding inequality constraints, are presented. These strategies include: a improved adaptive movement penalty method, a network topology based identification method and a limited times skipping method of trial iteration. The numerical results of several IEEE standard test system and two Chinese real network have shown that these improved strategies play a great role for underpinning the numerical stability, convergence characteristic and computation time saving.
出处
《中国电机工程学报》
EI
CSCD
北大核心
1999年第12期70-75,共6页
Proceedings of the CSEE
关键词
最优潮流
牛顿算法
电力系统
目标函数
optimal power flow
ill condition matrix
trial iteration
movement penalty
network topology
