摘要
柴油机的排气噪声中含有很多能反映柴油机故障的有用信息,因此从排气噪声中提取有效的特征向量可用于柴油机故障诊断。通过对柴油机不同状态下的排气噪声分析,对比了基于功率谱分析、小波包分解和AR模型的排气噪声特征提取方法,发现基于AR模型的排气噪声特征提取方法能有效地区分柴油机失火及气阀机构故障。
The exhaust noise signal of an engine contains many information which can reflect work state of the diesel engine. So extracting the useful eigenvector from the exhaust noise signal to diagnose the diesel engine is feasible. By analysing the exhaust noise signal under different conditions of diesel engine, feature extraction methods based on power spectrum analysis, wavelet package decomposition and auto regressive model were investigated. It was found that feature extraction method of auto regressive model can diagnose the misfire and valve train faults.
出处
《现代制造工程》
CSCD
北大核心
2012年第11期126-130,共5页
Modern Manufacturing Engineering
关键词
柴油机
排气噪声
功率谱
自回归模型
小波包
特征提取
故障诊断
diesel engine
exhaust noise
power spectrum
Auto Regressive （ AR ） model
wavelet package
feature extraction
fault diagnosis
