基于ROSETTA的柴油机故障诊断
首发时间:2010-04-16
摘要:本文介绍了粗糙集理论的核心内容和ROSETTA软件的特点,给出了基于ROSETTA的柴油机缸盖振动信号的故障诊断系统。以某型号大功率柴油机为例,首先将提取的缸盖振动信号经过小波包消噪和小波包能量谱分析,构造出用于故障诊断的特征值,然后应用ROSETTA软件约简特征属性,最后通过神经网络进行故障模式分类。通过对比ROSETTA软件处理前后神经网络的输出结果,表明小波包能量谱分析方法可以提取较好的特征值,而且粗糙集理论的特征属性约简,能有效地减少神经网络的输入节点数,提高故障分类的准确率。
关键词: 粗糙集 ROSETTA 小波包变换 RBF神经网络
For information in English, please click here
Diesel Engine Fault Diagnosis Based on ROSETTA
Abstract:This paper describes the core of rough sets theory and the ROSETTA software, it presents a diesel engine fault diagnosis system about cylinder head vibration signal. Take a high-power diesel engine for example, first use wavelet packet transform and energy spectrum analysis method analyze the signal, to take the eigenvalue for fault diagnosis. And then apply the ROSETTA software reduce the feature properties. Finally, distinguish between failure modes using neural network. By comparing the output of neural network based on ROSETTA, Show that the wavelet packet energy spectrum analysis method can get the good eigenvalue, and attributes reduction based on Rough Set, can effectively reduce the input of neural network nodes, which is to improve the accuracy of fault classification.
Keywords: Rough sets ROSETTA Wavelet packets transform RBF Artificial neural network
论文图表:
引用
No.4202552758512714****
同行评议
共计0人参与
勘误表
基于ROSETTA的柴油机故障诊断
评论
全部评论0/1000