| 基于MEA的粗糙集神经网络研究及应用 |
| |
| 引用本文： | 高金兰.基于MEA的粗糙集神经网络研究及应用[J].科学技术与工程,2011,11(23):5564-5568. |
| |
| 作者姓名： | 高金兰 |
| |
| 作者单位： | 1. 东北石油大学电气信息工程学院,大庆,163318
2. 燕山大学理学院,秦皇岛,066004 |
| |
| 摘 要： | 将思维进化算法、粗糙集和神经网络相结合,提出一种基于MEA的粗糙集神经网络,用于变压器故障诊断。此模型采用思维进化算法全局寻优的特点,搜索粗糙集属性约简离散断点的位置以及神经网络的连接权值和阈值,避免了常规粗糙集属性约简时复杂的手工试凑以及BP神经网络收敛速度慢、精度不高等缺点,有利于更快地收敛于全局最优解,提高系统的诊断速度和准确率。仿真结果表明了方法的有效性。
|
| 关 键 词： | 思维进化算法 粗糙集 属性约简 神经网络 变压器 故障诊断 |
| 收稿时间： | 2011/5/19 0:00:00 |
| 修稿时间： | 2011/5/19 0:00:00 |
Research and Application of Rough Set-BP Neural Network Based on MEA |
| |
| gaojinlan.Research and Application of Rough Set-BP Neural Network Based on MEA[J].Science Technology and Engineering,2011,11(23):5564-5568. |
| |
| Authors: | gaojinlan |
| |
| Institution: | GAO Jin-lan1,Gao Qian2 (College of Electrical & Information Engineering,Northeast Petroleum University1,Daqing 163318,P.R.China,College of Sciences,Yanshan University2,Qinhuangdao 066004,P.R.China) |
| |
| Abstract: | The mind evolutionary algorithm is combined,the rough set and the neural network,and a rough set-neural network based on MEA is proposed applying in transformer fault diagnosis.This model uses global optimization characteristics of the mind evolutionary algorithm to search rough set discrete breakpoints and neural network connection weights and thresholds,it avoids the conventional rough set complex handwork reduction and slow convergence and low precision of BP neural network,and benefits to find the globa... |
| |
| Keywords: | mind evolutionary algorithm rough set attribute reduction neural network transformer fault diagnosis |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
| 点击此处可从《科学技术与工程》浏览原始摘要信息 |
| 点击此处可从《科学技术与工程》下载免费的PDF全文 |
|
