基于小波包分析和神经网络柴油机故障诊断研究  

Study on Diesel Engine Fault Diagnosis Based on Wavelet Packet Analysis and Neural Networks

作　　者：赵龙庆[1] 王裕鹏[1] 

机构地区：[1]西南林学院交通机械与土木工程学院,云南昆明650224

出　　处：《小型内燃机与摩托车》2009年第3期9-11,18,共4页Small Internal Combustion Engine and Motorcycle

基　　金：云南省高校教学科研带头人资助项目(YJG2005028)

摘　　要：对于柴油机而言,及时的对其实施故障诊断是十分必要的。小波包分析能有效的对柴油机缸体振动信号进行消噪处理,并提取表征柴油机故障的振动信号能量特征向量作为BP神经网络的输入值。本文中,对4100QB柴油机缸盖振动信号实验数据采用单隐层BP网络进行训练,并对不同故障模式进行分类识别,实验验证表明这一方法是非常有效的。For diesel engine, it＇s necessary to carry out fault diagnosis in time. Wavelet packet analysis can de -noise the diesel engine cylinder head vibration signal effectively, and the eigenvector that can reflect the fault of diesel engine is extracted as the BP neural network input vector. In this article, the BP neural network is used to train the vibration signal experiment data obtained on the 4100QB diesel engine cylinder head, at the same time, the different fault mode is classified and recognized. Experimentation and validation show that the proposed method is very effective.

关 键 词：动力机械工程 小波包分析 BP神经网络 故障诊断 

分 类 号：TP206.3[自动化与计算机技术—检测技术与自动化装置] TK428[自动化与计算机技术—控制科学与工程]