摘要
以柴油机故障诊断为背景,研究了基于粗糙集理论的参数优化在故障诊断中的应用。首先采用小波包能量谱方法提取振动信号的特征参数,并用粗糙集理论对其进行属性约简,最后用RBF神经网络对各类故障进行辨识,结果表明:利用粗糙集约简后,通过减少神经网络的输入节点数,简化网络的结构,提高了诊断的准确率及效率。
Considering the diesel engine fault diagnosis,the rough set theory-based parameter optimization application in it was discussed.Firstly,the wavelet packet energy spectrum method was used to extract characteristic parameters of vibration signals,then the rough sets theory to conduct attribute reduction,finally,the RBF neural network to identify all kinds of faults.The results show that using rough set reduction can reduce the input nodes of neural network and simplify the network structure.This can improve the accuracy and efficiency of diagnosis.
出处
《化工自动化及仪表》
CAS
北大核心
2011年第1期40-43,共4页
Control and Instruments in Chemical Industry
基金
国家自然科学基金资助项目(50875247)
教育部博士点基金资助项目(20091420110002)
山西省自然科学基金资助项目(2007011070)
关键词
粗糙集
小波包变换
RBF神经网络
故障诊断
rough sets
wavelet packets transform
RBF artificial neural network
fault diagnosis
