摘要
变速箱产生的噪声主要是与转动频率相关的高频噪声,局域均值分解方法可以自适应地将信号分解成若干个调幅调频信号,即PF分量,这些调幅调频信号可以任意组合,突显某种信号特征进而起到滤波作用,从而实现信号的特征提取。文章针对于某变速箱2挡3000转运行时存在的异响主观判断认为其异响原因可能是变速箱内齿轮啮合存在故障,通过局域均值分解方法对变速箱噪声信号进行特征提取,并通过边际谱分析异响原因,确定噪声源。通过更换变速箱内故障齿轮后噪声消除,验证了该文所提出的研究方法是有效的。
Transmission Noise produced in mainly high frequency noise associated with the rotational frequency,Local Mean Decomposition method can adaptively decompose signal into a plurality of frequency-modulated and amplitude-modulated signal, namely the PF component, the signal can be any combination of features, highlighting some signal so as to filtering and realize the signal feature extraction. This paper aims at a certain transmission howling subjective judgment in block 2,3000 rpm operation,that its howling reason may be gear engaging fault in the transmission. Identify the noise source by the Local Mean Decomposition characteristics of the transmission signal, and spectrum analysis of howling reasons through the edge extraction. Confirmed the method is effective by replacing the fault gear in transmission after noise elimination.
出处
《汽车实用技术》
2015年第12期83-85,共3页
Automobile Applied Technology
关键词
局域均值分解
变速箱噪声
特征提取
Local Mean Decomposition
Transmission Noise
Local Mean Decomposition
