摘要
提出了一种小波变换和模糊C-均值聚类相结合的汽油机爆震诊断方法,并对某汽油机进行了爆震试验,利用小波变换方法从汽油机振动信号中提取了轻微爆震特征。以振动信号小波变换后的特征子带信号为样本信号,以特征子带信号的4个时域参数作为燃烧模式特征值,利用模糊C-均值聚类方法对燃烧模式进行分类,用贴近度方法进行爆震诊断。结果表明:小波结合模糊C-均值聚类方法能够利用振动信号有效地分类燃烧模式,并能诊断出轻微爆震。
A gasoline engine knock diagnostic method based on the algorithm combining wavelet transform with fuzzy C-means clustering was put forward.Knock experiments were carried out in a gasoline engine to extract light knock signatures from engine vibration signals by the wavelet transform method.Making use of sub-band signals generated by the wavelet transform of vibration signals as the sample signals,four time-domain parameters extracted from sub-band signals were confirmed as the evaluating indicators of combustion modes.Fuzzy C-means clustering method was used to classify combustion modes,and knock diagnosis was also carried out by using fuzzy neartude method.The result shows that the combustion modes of gasoline engine can be classified effectively from vibration signals by using wavelet transform and fuzzy C-means clustering method,and the slight knock combustion can also be diagnosed.
出处
《内燃机工程》
EI
CAS
CSCD
北大核心
2011年第4期56-59,64,共5页
Chinese Internal Combustion Engine Engineering
关键词
内燃机
汽油机
爆震诊断
小波变换
模糊C-均值聚类
IC engine
gasoline engine
knock diagnosis
wavelet transform
fuzzy C-means clustering
