| 基于模糊神经网络的电力系统预想事故排序方法 |
| |
| 引用本文： | 刘孟觉,余昆,刘皓明.基于模糊神经网络的电力系统预想事故排序方法[J].江苏电机工程,2007,26(3):1-3. |
| |
| 作者姓名： | 刘孟觉 余昆 刘皓明 |
| |
| 作者单位： | 1. 国电南瑞科技股份有限公司,江苏,南京,210008
2. 河海大学电气工程学院,江苏,南京,210098 |
| |
| 基金项目： | 高等学校博士学科点专项科研项目
,
河海大学校科研和教改项目 |
| |
| 摘 要： | 提出了一种基于模糊神经网络的电力系统预想事故排序新方法。该方法首先定义了反映预想事故严重程度的有功性能指标,同时构造3层人工神经网络(ANN)并采用误差反向传播(BP)算法加以训练;其次对ANN的输入用模糊神经网络进行特征选择,减少了输入层和中间隐含层的神经元个数及训练时间;最后通过IEEE 30节点系统验证了所提方法的有效性,仿真结果说明采用模糊神经网络进行输入量特征选择预处理可减少神经网络的训练时间。
|
| 关 键 词： | 预想事故排序 性能指标 特征选择 模糊神经网络 |
| 文章编号： | 1009-0665（2007）03-0001-03 |
| 收稿时间： | 2006-11-07 |
| 修稿时间： | 2007-01-29 |
Ranking Method of Potential Fault for Power Systems Based on Fuzzy Neural Networks |
| |
| LIU Meng-jue,YU Kun,LIU Hao-ming.Ranking Method of Potential Fault for Power Systems Based on Fuzzy Neural Networks[J].Jiangsu Electrical Engineering,2007,26(3):1-3. |
| |
| Authors: | LIU Meng-jue YU Kun LIU Hao-ming |
| |
| Affiliation: | 1. Nari Technology Development Limited Company,Nanjng 210008, China; 2. Hohai University, Nanjng 210098, China |
| |
| Abstract: | A new ranking method of potential fault for power systems based on fuzzy neural networks is developed.An active performance index is defined firstly which can reflect the severity degree of faults.And a three-layer ANN is built,which is trained using error back propagation algorithm.The fuzzy selection is used to reduce the neuron numbers between the input layer and the middle layer.The effectiveness of the proposed method is demonstrated on the IEEE 30 bus system.Results show that the proposed method can reduce the training time of neural networks. |
| |
| Keywords: | potential fault ranking performance index feature selection fuzzy neural networks |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
