摘要
本文利用熵测度理论分析了样本数据中每个辨识特征值的信息增益,然后再用神经网络的自适应共振理论模型和多层感知器处理这些特征值、辨识并恢复了不良数据,为电力系统状态估计提供了一个好的不良数据处理方法。
The Paper employs entropy measure theory to find out information profit of each recognition feature in sample data.Then,the adaptive resonance theory model and percepbons of artificial neural nets are used to identify and process bad data.A new method for bad data Processing in power system state estimation is thus built.
出处
《东北电力学院学报》
1995年第2期58-64,共7页
Journal of Northeast China Institute of Electric Power Engineering
关键词
不良数据
神经网络
电力系统
状态估计
bad data
artifical neural nets
adaptive resonance theory
perceptrons