混合噪声总烦恼度模型比较研究  

Comparative Study on Classical Models for Accumulation of Annoyance from Combined Noise

作　　者：闫靓[1] 陈克安[1] 王洁[1] 

机构地区：[1]西北工业大学航海学院环境工程系,西安710072

出　　处：《噪声与振动控制》2008年第4期149-154,共6页Noise and Vibration Control

基　　金：国家自然科学基金资助项目(批准号:10574104)

摘　　要：以城市环境中普遍存在且影响最大的混合噪声(由多种单一噪声以不同方式组合成的噪声)-道路交通噪声为例,在完成对典型车辆噪声分类的基础上,依据所提出的混合噪声组合方式,借助主观评价实验,重点研究3种不同类型车辆噪声"同时作用"下的混合噪声烦恼度,对现有总烦恼度模型进行全面比较分析。研究发现:(1)矢量叠加模型的计算值与评价实验测量值之间的相关性最好,主导源模型次之,线性回归模型较差;(2)主导源模型的预测误差最小,线性回归模型较大,矢量叠加模型预测误差最大。由此可见,混合噪声总烦恼度模型评价性能的优劣不能仅依据模型预测值和评价测量数据的相关性高低来衡量。Generally, the community noise which constituted in different way by single noise originating from different types of source can be called combined noise. A laboratory study was conducted to investigate the predicted performances of various classical models for total noise annoyance based on the known rating of annoyance caused by a selection of various individual vehicles. The results of the quality evaluation indicate that： （1） the vector summation model shows a higher performance in predicting the outcome of subjective experiments, next one is the strongest component model, the linear regression model is inferior; （2） the investigation carried out by two measures of model performance has revealed that the strongest component model is the best, then is the linear regression model, the vector summation model is the worst. Therefore, the correlation coefficient between models results and （experiment or survey） data is not a credible quality measure for models.

关 键 词：声学 混合噪声 烦恼度 语义细分法 

分 类 号：TB529[理学—物理]