基于小波变换的非平稳排气噪声信号阶次分析方法  

Ordering analysis of non-stationary exhaust noise signals based on wavelet transform

作　　者：刘海涛 LIU Haitao(School of Mechanotronics and Vehicle Engineering,East China Jiaotong University,Nanchang 330013,China)

机构地区：[1]华东交通大学机电与车辆工程学院

出　　处：《振动与冲击》2019年第22期29-35,51,共8页Journal of Vibration and Shock

基　　金：国家自然科学基金(51765017);江西省交通运输厅科技项目(2015D0062);江西省自然科学基金(20161BAB216136);华东交通大学科研基础经费(26541022)

摘　　要：汽车排气阶次噪声提取对于汽车声品质以及汽车分类识别具有重要的意义。为准确提取阶次噪声的时域信号,提出一种基于小波变换的非线性多分辨率的细化分析计算方法。通过理想带通传递函数构建基小波函数,基小波函数通过平移和伸缩实现信号的局部细化分析。探索小波函数的截取方式对频谱泄漏的影响,选取合适的窗函数截取小波函数,并通过与时域噪声信号的相关变换准确提取各阶次成分的时域波动信号。通过实车测试获取加速工况下的转速脉冲信号和排气辐射噪声信号,对以上方法进行验证。结果表明,提出的分析方法可准确提取出非平稳排气噪声中的阶次成分,为排气噪声阶次分析提供标准可靠的信号处理手段。The noise ordering analysis of vehicle exhaust is of great significance for vehicle sound quality inspection as well as for vehicle classification and recognition.In order to extract the time domain signals of the order components of exhaust noise,a nonlinear refining analysis method with multi-resolution based on wavelet transform was proposed.The basis wavelet function was constructed by using the ideal band-pass transfer function.The local refinement analysis of signals was realized by translating and scaling the basis wavelet function.The influence of the interception approach for the wavelet function on the spectrum leakage was explored and a proper window function was selected to intercept the wavelet function,which was used to accurately extract the time domain signal of each order component through the correlation transform on the time domain noise signal.At last,the speed pulse signal and the exhaust noise signals under accelerating working condition were obtained by real vehicle tests and the analysis method proposed was verified.The results show that the analysis method proposed can accurately extract the order components of the non-stationary exhaust noise signals,thus providing a reliable and standard signal processing method for the ordering analysis of exhaust noises.

关 键 词：非平稳时变信号 阶次 基小波函数 窗函数 相关变换 

分 类 号：TN911.72[电子电信—通信与信息系统] TB52[电子电信—信息与通信工程]