基于自组织竞争神经网络的电力系统暂态稳定研究  

Study of Power System Transient Stability Based on Self-organizing Competition Artificial Neural Network

作　　者：危雪[1] 周云海[1] 胡翔勇[1] 

机构地区：[1]三峡大学电气信息学院,湖北宜昌443002

出　　处：《三峡大学学报（自然科学版）》2008年第5期25-28,共4页Journal of China Three Gorges University：Natural Sciences

摘　　要：大停电事故和电力需求紧张,使得大扰动下电力系统的暂态稳定问题日显突出.如何能快速判断电力系统的暂态稳定性能,是研究的重点.文章以综合性能指标法中的性能指标作为反映电力系统暂态稳定性能的特征指标,设计自组织竞争人工神经网络分类器进行暂态稳定问题的分类判断.通过算例验证了该分类器可将暂态稳定问题分为稳定、不稳定和模糊类3种类型,有利于对稳定性能进行快速判断,从而减少了待详细分析的事故数目,而无需利用全过程时域仿真计算,有利于进行在线的动态安全评估.Power system transient stability problem becomes more important nowadays. Six indexes of integrated performance index method are used to design a self-organizing ANN classifier to estimate transient stability problem. This kind of classifier can distinguish stable accident and unstable accident except for several illegible parts. It has better classified effect, so can help us to estimate transient problem quickly by reducing the number of accidents that need to be analyzed in detail. Furthermore, it is unnecessary to have complete calculation by time-domain simulation, so as to benefit the online dynamic security assessment.

关 键 词：电力系统 暂态稳定性能 自组织竞争人工神经网络 快速判断 

分 类 号：TM73[电气工程—电力系统及自动化]