摘要
结合人工神经网络（ＡＮＮ）和短时数字仿真提出一个用于在线暂态安全评估的事故筛选方法，将３层ＢＰ网络作为模式分类器，用来建立稳定评估结果和所选特征量之间的映射关系。在故障切除时刻终止的短时数字仿真被用来生成ＡＮＮ的输入量，每个ＡＮＮ处理一个特定的事故状态。使用一个半监督学习算法，ＡＮＮ可产生一个能够指示相对稳定度的连续分布的暂态稳定指标。基于这个连续分布的稳定指标，设置一个相对保守的分类门槛值，避免了不安全状态的漏报。１０机新英格兰电力系统的应用结果证实了该方法的有效性。
Integration of short-duration numerical simulation techniques and artificial neural networks is investigated for application to contingency screening of dynamic security assessment. In the proposed approach, the back-propagation neural networks are employed to assess transient stability of power systems. The short-duration numerical simulation is employed to produce the input atributes to the ANNs. Each ANN is designed to handle a single contingency scenario. The ANN can derive a continuous-spread stability index to indicate the relative stability degree by means of a semi-supervised learning algorithm. Based on the stability index, a conservative classification threshold is set to avoid omission of insecure cases. Applications to the 10-unit New England Power System demonstrate the validity of the proposed approach.
出处
《电力系统自动化》
EI
CSCD
北大核心
1999年第8期16-19,26,共5页
Automation of Electric Power Systems
关键词
电力系统
暂态稳定
神经网络
安全评估
事故筛选
power systems transient stability neural networks security assessment contingency screening
