摘要
通过对发电机运动方程理论分析,推导和仿真实验证明了最佳重合时刻系统稳定性判据的合理性.借鉴基于模拟植物生长的改进BP神经网络最佳重合时刻智能捕捉算法,利用经验模态分解提取样本特征值作为网络的输入,将通过系统稳定性判据确定的最佳重合时刻作为网络的期望输出进行网络训练.仿真实验证明,该算法测试误差非常小,且捕捉到的最佳重合时刻可以使系统更快稳定.
By analyzing generator motion equations,the rationality of system stability criterion at the optimal reclosing time is deducted.The criterion is proved to be reasonable by simulation experiments.An intelligent capture algorithm of the optimal reclosing time is proposed based on a BP neural network improved over plant growth simulations.The neural network is trained by using the Empirical Mode Decomposition to extract sample characteristics as the network input and the optimal reclosing time to be determined by the system stability criterion as the desired output of the network.Simulation experiments show that the algorithm has high accuracy and the optimal reclosing time captured can make the system run faster and more stable.
作者
杜端强
李春明
李智
DU Duan-qiang;LI Chun-ming;LI Zhi(Urumqi Power Distribution Company of State Grid Xinjiang Electric Power Co.Ltd.Urumqi 830011,China;College of Information Engineering,Inner Mongolia University of Technology,Hohhot 010080,China)
出处
《内蒙古工业大学学报(自然科学版)》
2019年第5期378-383,共6页
Journal of Inner Mongolia University of Technology:Natural Science Edition
关键词
最佳重合时刻
系统稳定性判据
模拟植物生长算法
BP神经网络
optimal reclosing time
system stability criterion
plant growth simulation algorithm
BP neural network