基于粗糙集的特征值优化及柴油机故障诊断  

Characteristic Optimization and Diesel Engine Fault Diagnosis Based on Rough Set

作　　者：田静宜[1] 潘宏侠[1] 杨业[1] 

机构地区：[1]中北大学机械工程与自动化学院,山西太原030051

出　　处：《车用发动机》2010年第5期84-89,共6页Vehicle Engine

基　　金：国家自然科学基金资助(50875247);教育部博士点基金资助(20091420110002);山西省自然科学基金资助(2007011070)

摘　　要：为提高故障诊断的效率,给出了一种基于粗糙集理论的柴油机故障诊断系统。以某大功率柴油机为例,采用时域频域分析和小波包能量谱分析两种方法提取特征值,通过对比优选,将敏感性和稳定性较好的小波包能量谱特征值应用粗糙集理论进行优化,最后通过神经网络进行故障模式分类。试验表明,小波包能量谱分析方法可以提取敏感性和稳定性较好的特征值,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。To improve the efficiency of fault diagnosis,a fault diagnosis system based on rough set was put forward.For a high-power diesel engine,the fault characteristics were extracted with the time-frequency analysis and wavelet packet energy spectrum analysis method.By comparison and analysis,it was decided that the latter extracted characteristics,which had better sensitivity and stability,were optimized with rough set.Finally,the fault modes were categorized with neural network.The results show that the characteristics extracted by wavelet packet energy spectrum method have better sensitivity and stability.With rough set,the characteristic attributes are simplified,which reduces the input nodes of neural network.Accordingly,the accuracy of fault classification improves.

关 键 词：柴油机 故障诊断 粗糙集 小波包变换 神经网络 

分 类 号：TK421[动力工程及工程热物理—动力机械及工程] TP182[自动化与计算机技术—控制理论与控制工程]