粗糙集——神经网络故障诊断方法研究
Research on Rough Set-Neural Network Fault Diagnosis Method
粗糙集理论具有处理不完整样本数据的优点，本文提出了使用粗糙集理论优化神经网络故障诊断系统的算法。在保证故障分类结果基本不变的情况下，利用粗糙集理论对原始故障特征信息表进行约简,得到最简决策表,导出诊断规则, 再输入神经网络进行训练学习。通过一个具体实例分析表明，该方法有效地减少了输入层神经元的个数，提高了网络的学习速率和诊断的准确率，在故障诊断中有良好的应用前景。
Rough set theory (RST in short) can deal with incomplete data, this paper propose the algorithm of using rough set theory to optimize neural network in fault diagnosis system. Not changing classification ability basically, RST is used to simplify the original fault information sheet, get the simplest expert diagnosis rules, and then enter the neural network for training and learning. At last, through a concrete example, the experiment result shows that this method can effectively reduce the input neurons number and improve the network\\\\\\\
谢刚、师秀川
自动化技术经济自动化技术、自动化技术设备
粗糙集故障诊断差别矩阵神经网络
Rough SetFault DiagnosisDiscernibility MatrixNeural NetworkExpert System
谢刚,师秀川.粗糙集——神经网络故障诊断方法研究[EB/OL].(2008-12-02)[2025-08-30].http://www.paper.edu.cn/releasepaper/content/200812-61.点此复制
评论