| 概率神经网络故障诊断的粗糙集优化方法 |
| |
| 引用本文： | 徐袭,李浩军,范学鑫. 概率神经网络故障诊断的粗糙集优化方法[J]. 微计算机信息, 2007, 23(19): 202-204 |
| |
| 作者姓名： | 徐袭 李浩军 范学鑫 |
| |
| 作者单位： | 1. 430033,湖北武汉,海军工程大学,电气与信息工程学院
2. 116000,辽宁大连,91278部队50分队 |
| |
| 基金项目： | 国家自然科学基金委员会创新研究群体科学基金资助（50421703） |
| |
| 摘 要： | 针对径向基概率神经网络故障诊断输入量过多会影响网络学习效率的问题,提出了一种基于粗糙集的概率神经网络故障诊断优化方法.该方法用模糊C均值聚类将故障诊断训练数据离散化,使用粗糙集化简由输入输出属性构成的决策表,利用约简后的输入量重新构建神经网络故障诊断模型,使原有模型得到优化并以柴油机故障诊断为例说明该方法的有效性.
|
| 关 键 词： | 粗糙集 模糊C均值聚类 概率神经网络 故障诊断 |
| 文章编号： | 1008-0570（2007）07-1-0202-03 |
| 修稿时间： | 2007-05-132007-06-15 |
Optimizing Method with Rough Set for Probabilistic Neural Network Fault Diagnosis |
| |
| XU XI,LI HAOJUN,FAN XUEXIN. Optimizing Method with Rough Set for Probabilistic Neural Network Fault Diagnosis[J]. Control & Automation, 2007, 23(19): 202-204 |
| |
| Authors: | XU XI LI HAOJUN FAN XUEXIN |
| |
| Abstract: | A kind of optimizing method based on rough set is proposed for problem of network learn efficiency because of too much input values for probabilistic neural network fault diagnosis. The method use FCM clustering to disperse the train data of fault diag- nosis and then to predigest the decision table constructed by I/O attributes with rough set. The inhere model of NN fault diagnosis model can optimize by Input value after reduced reconstructed the new model. The method is validity through the example of diesel engine fault diagnosis. |
| |
| Keywords: | Rough set Fuzzy C- Means clustering Probabilistic neural network Fault diagnosis |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
