低频噪声声品质及与频谱特征和烦恼度关系的研究
文献类型：学位论文
| 作者 | 孙艳 |
| 学位类别 | 博士 |
| 答辩日期 | 2009-05-31 |
| 授予单位 | 中国科学院声学研究所 |
| 授予地点 | 声学研究所 |
| 关键词 | 低频噪声 频谱特征 声品质 正交实验 烦恼度 |
| 其他题名 | Investigation of low-frequency sound quality and relations between sound quality and spectrum characteristics and annoyance |
| 学位专业 | 声学 |
| 中文摘要 | 随着社会的不断进步，人们对声环境的要求越来越高。因为人对低频噪声的主观反应高于同一响度条件下的其他噪声，它的影响越来越受到关注。如何描述低频噪声所独有的特征，这些特征对人产生何种影响，怎样建立噪声信号特征量模型与主观感觉之间的关系等，已经成为低频噪声研究的热点。 本文运用正交实验方法，系统地研究了有关低频噪声中频谱平衡、信号调制等因素对心理声学模型和主观反应的影响，主要工作如下： （1） 详细论述了低频噪声信号中四个频谱特征：不同频率下降点，不同下降斜率，不同频率调制因子和用于调制信号的窄带噪声不同截止频率对心理声学模型和主观感觉的影响。改变了目前声品质研究中只是针对某一类特定声信号建立烦恼度预测模型，而没有进一步系统研究频谱特征与主观反应之间关系的现状。 （2） 采用正交实验方法，详细分析了现有心理声学参量在描述本文所讨论信号频谱特征时的有效性和局限性。研究发现，现有的尖锐度和主观低沉度没有相关性，而且不同频率成份对低沉度的影响程度也不一样。粗糙度和有调度也不能完全表征相应的主观感觉，粗糙感不但随信号声压级起伏而变化，还与信号的峰值形状有关，峰值宽度越宽，粗糙感越强。对于窄带噪声调制的声信号，仍可以用掩蔽深度来描述波动强度的大小：掩蔽深度越大，波动强度越大。 （3） 修正了现有的有调度模型。针对所研究的声信号，实验分析结果显示对有调感影响最大的因素是有调音部分的声压级大小和谱峰形状，据此本文通过主观评价实验对现有模型中的带宽计权因子进行了修正，从而增强了该因子计算模型的鲁棒性，提高了有调度模型的准确性。 （4） 建立了烦恼度预测模型。文中首先采用正交实验分析结果，确定对主观感觉有显著影响的频谱特征，然后选择与影响主观感觉一致的频谱特征所影响的心理声学模型，通过多元线性回归方法建立烦恼度预测模型，为从大量计算模型中筛选相互独立、冗余性小的特征量提出新的解决思路，有效的提高了评价研究工作的效率。通过对预测模型的分析得到，对于含有较高有调音的噪声有调度对烦恼度的影响比较明显，但是响度仍起主导作用。 |
| 英文摘要 | The influence of low-frequency noise has obtained an increasing attention because adverse effects of low-frequency noise are more serious than the other components of the noise at the same sound pressure levels. How to describe the special characters of low-frequency noise? What kind of characteristics has a corresponding impact on human beings? How to establish the relationship between the models of the characteristic noise signals and the subjective sensations? These have become academic hot topics of low-frequency noise. In this thesis, an orthogonal array analysis method is applied to the study of sound quality in order to systematicly investigate the influence of frequency spectrum characteristics of low-frequency noise on subjective sensations about frequency spectrum balance and narrow band frequency modulation. The main contents are as follows： （1） A detailed description of the spectrum characteristics effect on the subjective sensations is presented, including different drop points and drop slope in the frequency spectrum, different ranges of frequency modulation and different cut-off frequencies. （2） Used the orthogonal array analysis method validities and limities of some psychoacoustic parameters have been discussed. Different frequencies will have various contributions to the low-frequency index apart from the power ratio of the high frequencies to the low frequencies. Therefore, when curve fitting is used, sharpness and roughness are considered to describe signal characteristics. Tonality can well define roughness of the noise which carries high tones .The peak power spectrum width of the signal will affect the subjective roughness severely besides the fluctuation of the noise. The wider the peak width is, the stronger roughness is. Moreover, masking depth might still be used to describe fluctuation sensations about the sound signal. The deeper the masking depth, the severer fluctuation is. （3） Then the model of tonality is modified. According to the analysis of subjective assessment, a predominant influence on the tonality is the sound pressure level and the shape of peak spectrum of tones. Based on this result, the factor of peak spectrum width is determined to improve the robustness of computation and the accuracy of the model. （4） Finally the predictive model of annoyance is built based on a fixed model and the orthogonal array analysis. Results are also used to extract less information redundancy and salient interpretive features from a large quantity of parameters at a condition of a few samples. |
| 语种 | 中文 |
| 公开日期 | 2011-05-07 |
| 页码 | 129 |
| 源URL | [http://159.226.59.140/handle/311008/154]  |
| 专题 | 声学研究所_声学所博硕士学位论文_1981-2009博硕士学位论文 |
| 推荐引用方式 GB/T 7714 | 孙艳. 低频噪声声品质及与频谱特征和烦恼度关系的研究[D]. 声学研究所. 中国科学院声学研究所. 2009. |
入库方式： OAI收割
来源：声学研究所
浏览0
下载0
收藏0
其他版本
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。