中文摘要：

目前，噪声识别及特征分析主要是通过信号处理及变换技术实现的，对噪声频率结构的分析是研究噪声机理的重要手段。针对噪声的常规信号分析中对噪声的主观性考虑不足的缺陷，以汽车车内稳速噪声及阶次扫频噪声两大类型噪声信号类型为分析对象，综合应用基于人耳主观特性的计权声级、特征响度模型等分析方法对上述类型噪声进行了分析对比。结果表明，采用计权声级、特征响度等方法（特别是特征响度分析法）能更好地进行噪声识别和特性分析，从而得到与主观相一致的结论。

英文摘要：

Currently, noise identification and analysis are mainly realized by signal processing and transformation technique, the frequency structure analysis of noise is an important means to research noise mechanism. This paper discussed the shortcomings of common signal analysis m.ethod, such as the lack of subjectivity consideration for noise and so on. The weighting sound level and specific loudness model which based on human ear subjective characteristics were applied to analyze and compare the two typical noise such as steady vehicle interior noise and order-swept noise. The results show that the weighting sound level and specific loudness model （especially the loudness method） can be very helpful to noise source identification and it＇s mechanism analysis, and the results are more consistent to subjective characteristic.