基于小波变换的柴油机故障智能检测新方法  

Intelligent Fault Detection Method for Diesel Engines Based on Wavelet Transform

作　　者：姜巍[1] 张卫宁[1] 

机构地区：[1]山东大学信息科学与工程学院,山东济南250100

出　　处：《控制工程》2005年第3期277-280,共4页Control Engineering of China

摘　　要：提出了一种智能检测柴油机喷油压力信号特征点的新方法,可以对柴油机一些常见故障进行不停机的检测。通过对油压信号特征的分析,对之进行了连续小波变换,根据油压信号在不同阶段的频率特征选定相应尺度的小波系数,然后利用这些小波系数的模极小值进行特征点的检测。将现场实测信号在Matlab中进行仿真,建立了喷油系统正常工作时的模板向量。实验证明该方法的准确率高,而且便于在各种数字信号处理终端中实现,是有效可行的。The feature points of diesel engine fuel injection pressure signals is introduced to detect some common faults with the engine running.Continuous wavelet transform is set to the fuel injection pressure signals.Some very scales are selected by the frequency character of the signals,and then wavelet coefficient modulus minima are used in the detection of these points.Through the Matlab simulation of experimental data,standard template vectors are built when fuel injection system running in gear.It experimentally proves that this method can get high precision.In addition,it can be implemented in many digital signal processing terminals conveniently and it is effective and feasible.

关 键 词：智能检测 柴油机故障 连续小波变换 数字信号处理 小波系数 MADAB 压力信号 常见故障 信号特征 频率特征 现场实测 喷油系统 实验证明 特征点 不停机 极小值 工作时 准确率 油压 仿真 

分 类 号：TP274[自动化与计算机技术—检测技术与自动化装置] U664.121[自动化与计算机技术—控制科学与工程]