车内声品质主客观评价分析

采集某纯电动汽车车内噪声样本,作为为实验评价对象,选取的车内噪声主观评价方法为等级评分法,并对噪声样本的响度、粗糙度、语义清晰度和尖锐度,此4种心理声学客观参量进行计算,通过相关性分析,得出各评价人员间的相关系数和主观评价值和心理声学客观参量间的相关性,排除相关性低的主观评价人员和心理声学客观参量,以主观评价值为因变量,以心理声学客观参量为自变量。建立车内声品质多元线性回归预测模型,预测车内声品质情况,以节省大量人力和物力成本,实现声品质的快速评价。     

听觉频带声能量对车内声品质客观评价影响研究

响度、尖锐度等心理声学客观参量常作为声学特征建立车辆噪声品质客观评价模型。考虑人耳听觉多频带滤波特性,研究听觉频带声能量对声品质客观评价的影响。建立基于频带声能量的汽车车内噪声品质客观量化模型,提高声品质识别的精度和稳定性。采集汽车匀速车内噪声并进行综合烦躁度主观评价试验,建立小波包耳蜗滤波器组提取听觉特征频带内声能量,利用支持向量机建立频带能量与主观评价结果之间的回归模型。交叉验证结果表明,相比于建立的基于心理声学客观参量综合烦躁度模型,建立的人耳听觉特征频带能量综合烦躁度客观评价模型预测的误差均值和误差方差更小,车内噪声品质评价的精度、稳定性均有提高。     

基于支持向量机的车内噪声声品质预测

提出了采用支持向量机的车内噪声声品质预测方法,对采集的车内噪声样本采用基于小样本理论的支持向量机回归方法,建立车内噪声声品质客观评价参量与主观评价结果的关系模型对车内噪声声品质进行预测.实例分析表明,选取适当的车内噪声声品质客观评价参量,利用支持向量机回归方法建立的车内噪声声品质预测模型的预测精度较高.     

基于声-固耦合模型的车内噪声响度参数的预测与分析研究

针对传统A计权声压级评价指标对噪声低频成分衰减较大,常常出现车内声压级达标,声品质不合格的问题。引入心理声学参数的响度参量,利用虚拟仿真技术分析评价车内声学特性。建立驾驶室声-固耦合有限元模型,结合试验激励数据,进行基于模态的声学响应计算。在Matlab平台上,建立车内声品质客观心理声学参数响度的计算模型,对比预测驾驶室内场点的声压级和响度分布,结合声压级和响度结构板块贡献量分析,研究声压级和响度参量评价驾驶室结构特性的差别并识别驾驶室主要噪声源。以此为基础指导优化驾驶室结构阻尼铺设位置,综合提高车内声学品质。     

基于降维-支持向量回归的车内稳态声品质预测

基于信号分析与机器学习方法,提出基于降维-支持向量回归(Dimension Reduction-Support Vector Regression,DRSVR)的声品质主观预测模型。以车内稳态声样本为研究对象,计算并分析了其基本物理参数、心理声学参数。运用成对比较法对声样本进行了主观偏好性实验,验证了烦恼度(Psychoacoustic Annoyance,PA)模型初步判断该类样本声品质好坏的可用性。通过因子分析、聚类分析与相关分析,完成了声样本的降维,提取出了主要影响参量,再以支持向量回归的方法建立了主观评价预测模型。相关分析显示,基于降维-支持向量回归预测模型的计算值与主观评价值的相关性较高,其预测能力比未降维的支持向量回归模型更优,证明运用DR-SVR方法对车内稳态声品质预测是有效的。     

基于心理声学的高速列车车内噪声预测及选材方法

为改善高速列车车内声场环境,优选出最佳的车体结构选材方案,以高速列车车顶结构为研究对象,将车顶隔声性能、客室内噪声以及心理声学三者联系起来,提出了基于心理声学的高速列车车内噪声预测及选材方法,并得出结论:基于心理声学客观参量的车内噪声评价方法相比于A声级能够更为合理地体现人耳的主观感受,并以此为基础确定出最佳的车顶结构选材方案;将车顶结构中的玻璃棉替换成岩棉后,结构计权隔声量提高1.3 d B,车内A计权声压级总值差异只有0.01 d BA,总响度和总噪度却发生恶化分别提高了1~1.5 sone和1~2 Noy;将5 mm隔音垫调至靠近车外一侧,结构计权隔声量提高1.4 d B,而车内A计权声压级总值车内A计权声压级总值差异0.1 d BA左右,而总响度、总噪度总值也发生恶化分别提高了3~5 sone和2~3 Noy。所提出的研究和评价方法对于高速列车车内噪声预测及车体结构的选材具有工程实用性和借鉴意义。     

声品质主观评价方法对比分析

声品质主观评价方法有:排序法、等级评分法、成对比较法及参考语义细分法等。根据汽车车内噪声测量标准设计噪声采集方案,对平稳和非平稳车内噪声进行了数据采集、信号预处理。分别采用四种主观评价分法对车内噪声的烦躁度进行了主观评价试验,将试验结果统计分析与客观参数值相关性分析以及对比分析和理论分析研究。结果表明,每种主观评价方法都有其各自的适用性,综合考虑参考语义细分法适用性强和应用广泛,更适合车内噪声声品质的主观评价。     

基于噪声主动控制方法的非稳态车内声品质优化研究

首先对16辆各种型号的轿车进行了加速条件下车内噪声样本的采集,通过噪声信号的主客观分析研究,得出主观评价结果与客观物理参量之间的相关性,明确了加速状态下响度、尖锐度及粗糙度是影响车内声品质的主要因素。而后,对声品质最差的7号车运用噪声主动控制方法进行车内噪声优化实验,通过对比发现,车内噪声的响度、尖锐度及粗糙度都有明显的减少。将数据代入声品质客观计算模型,同时进行主观评价实验,结果表明其评分等级从16级降至12级,从而证明了噪声主动控制方法能有效提高加速条件下车内噪声的声品质。     

基于LS-SVM算法的加速车内噪声品质评价模型

以汽车加速车内噪声为研究对象,采用最小二乘支持向量机(LS-SVM)方法建立了声品质评价模型。分别以车辆噪声的客观评价结果和主观烦躁度作为模型的输入和输出,在相关分析和显著性检验的基础上,以响度、尖锐度、粗糙度、清晰度指数和A声级为变量建立了LS-SVM声品质评价模型。对未知噪声样本的预测检验表明:预测结果与主观烦躁度具有很高的相关性,预测精度高于多元线性回归方法。所建立的模型具有良好的泛化能力,可用于加速车内噪声品质的预测。     

乘用车车内噪声预测与排除技术浅析

人们对声音的主观感觉总是敏感而模糊的,随着人们对声音的不断研究,逐渐形成了描述声音特性的心理声学客观参数,如响度、尖锐度、粗糙度等。本文对车内噪声预测与排除技术进行了理论浅析。以响度、尖锐度和清晰度指数为输入变量,综合满意度为输出变量,研究了基于多元线性回归方法的车内噪声品质预测,建立了预测模型,并对模型进行了验证,结果显示,模型较为准确。     

Short-term annoyance reduction evoked by active noise control method on turboprop regional aircraft cabin noise

Since the 2000s, the advancing global economy has increased the demand for turboprop regional aircraft, and various studies have been conducted to resolve aircraft noise-related problems. Despite the progress made by previous studies, turboprop regional aircraft still have a poor noise environment compared to other vehicles. In addition, although an application of active noise control (ANC) method had improved the noise environment of the cabin, none of studies were conducted in consideration of a relationship between the application of ANC method and the resulting change of an (short-term) annoyance. Therefore, in the current study, an annoyance reduction evoked by active noise control (ANC) methodology on a turboprop regional aircraft cabin noise was investigated in three experiments. In the first experiment, 50 participants estimated the perception limit of psychoacoustic parameters on aircraft interior noise to identify the cognitive characteristic of the noise. Furthermore, the noise reduction achieved by applying ANC was identified based on the empirical results of the second experiment. Finally, the third experiment examined the relationship between psychoacoustic parameters (loudness and sharpness) and annoyance with the assistance of 50 participants. As a result, the difference between the noise reduction and the annoyance reduction was identified. Furthermore, the degree of annoyance reduction of implementing ANC in cabin noise was identified.

     

Comparative study on civil aircraft noise metrics as annoyance estimators for interoperability between other aircraft noise metrics

In Korea, the national noise evaluation system for transportation noise was divided into two kinds of noise metrics: WECPNL (weighted equivalent continuous perceived noise level) for aircraft noise and L _Aeq for other noises. For administrative and political application of these metrics, a trade-off between them must be maintained. Traditionally, conversion between noise metrics has included correlations between the two metrics based on measured physical quantity. However, conversion between two noise metrics should consider the subjective or emotional aspects of people exposed to noise. In the present study, a field study was conducted in 11 sites in the vicinity of Gimpo airport to propose a rational method to convert Korean WECPNL into DENL (day-evening-night average sound level). Annoyance models for aircraft noise in terms of Korean WECPNL and DENL were established by linear regression analyses for 382 useful data. CBA (conversion based on noise annoyance) was proposed by using the two annoyance models. The validity of the proposed equivalent annoyance method was demonstrated by comparing among dose-response relationships in terms of measured DENL, DENL estimated by the CBA, and DENL estimated by the three empirical relations between WECPNL and DENL.     

Design of new sound metric and its application for quantification of an axle gear whine sound by utilizing artificial neural network

The gear whine sound of an axle system is one of the most important sound qualities in a sport utility vehicle (SUV). Previous work has shown that, because of masking effects, it is difficult to evaluate the gear whine sound objectively by using only the A-weighted sound pressure level. In this paper, a new objective evaluation method for this sound was developed by using new sound metrics, which are developed based on the increment of signal to noise ration and the psychoacoustic parameters in the paper, and the artificial neural network (ANN) used for the modeling of the correlation between objective and subjective evaluation. This model developed by using ANN was applied to the objective evaluation of the axle-gear whine sound for real SUVs and the output of the model was compared with subjective evaluation. The results indicate a good correlation of over 90 percent between the subjective and objective evaluations. This paper was recommended for publication in revised form by Associate Editor Yeon June Kang Professor Sang-Kwon Lee received a Ph.D. degree in ISVR (Institute of Sound and Vibration Research) from Southampton University in 1998. He joined Hyundai Motor Research Center in Korea, working with the Automotive Noise and Vibration Control Group from 1985 to 1994. He has been the Professor at the Department of Mechanical Engineering, Inha University, Inchon, Korea, since March 1999. His research interests are the digital signal processing, NVH (noise vibration harahness), condition monitoring, product sound quality design and active control.     

Psychoacoustic study on contribution of fan noise to engine noise

There are more researches on engine fan noise control focusing on reducing fan noise level through optimizing fan structure,and a lot of research achievements have been obtained.However,researches on the effect of fan noise to engine noise quality are lacking.The influences of the effects of fan structure optimization on the engine noise quality are unclear.Thus,there will be a decline in fan noise level,but the deterioration of engine noise quality.Aiming at the above problems,in consideration of fan structure design and engine noise quality,an innovative method to analyze the contribution of fan noise to engine noise quality using psychoacoustic theory is proposed.The noises of diesel engine installing different cooling fans are measured by using the acoustic pressure method.The experiment results are regarded as analysis samples.The model of sensory pleasantness is used to analyze the sound quality of a diesel engine with different cooling fans.Results show that after installing 10-blade fan in medium diameter the sensory pleasantness at each test point is increased,and the increase is 13.53% on average,which indicate the improvement of the engine noise quality.In order to verify the psychoacoustical analysis,the subjective assessment is carried out.The test result shows the noise quality of engine installed10-blade fan in medium diameter is most superior.1/3 octave frequency spectrum analysis is used to study the reason of the improvement of engine noise quality.It is found that after installing proper cooling fan the sound pressure level below 400 Hz are obviously increased,the frequency assignment and spectral envelope are more reasonable and a proper cooling fan can optimize the spectrum structure of the engine noise.The psychoacoustic study is applied in the contribution of fan noise to engine noise,and the idea of engine sound quality improvement through the structure optimization is proposed.     

汽车声品质的GA-BP网络预测与权重分析

为了高效而准确地评价与控制车内噪声品质,以B级车稳态工况下副驾位置的车内噪声为研究对象,采用等级评分法对采集到的声音样本进行了主观评价试验,同时计算了7个客观参数。以客观参量为输入,声品质主观结果为输出,引入基于遗传算法的BP神经网络建立了声品质预测模型。实验显示该模型输出结果与实际评分的相关系数达到0.928,检验组的预测最大误差为±8%。以所建模型的连接权值,分析了客观参数对主观评价结果的贡献度,并以影响系数较大的参数为输入重新构建了预测模型。研究结果表明：稳态工况下,车内声品质主要受响度、粗糙度和尖锐度的影响,其预测模型可由这3个参数来描述。     

Annoyance rate evaluation method on ride comfort of vehicle suspension system

The existing researches of the evaluation method of ride comfort of vehicle mainly focus on the level of human feelings to vibration. The level of human feelings to vibration is influenced by many factors, however, the ride comfort according to the common principle of probability and statistics and simple binary logic is tmable to reflect these uncertainties. The random fuzzy evaluation model from people subjective response to vibration is adopted in the paper, these uncertainties are analyzed from the angle of psychological physics. Discussing the traditional evaluation of ride comfort during vehicle vibration, a fuzzily random evaluation model on the basis of annoyance rate is proposed for the human body＇s subjective response to vibration, with relevant fuzzy membership function and probability distribution given. A half-car four degrees of freedom suspension vibration model is described, subject to irregular excitations from the road surface, with the aid of software Matlab/Simulink. A new kind of evaluation method for ride comfort of vehicles is proposed in the paper, i.e., the annoyance rate evaluation method. The genetic algorithm and neural network control theory are used to control the system. Simulation results are obtained, such as the comparison of comfort reaction to vibration environments between before and after control, relationship of annoyance rate to vibration frequency and weighted acceleration, based on ISO 2631 / 1 （1982）, ISO 2631-1 （1997） and annoyance rate evaluation method, respectively. Simulated assessment results indicate that the proposed active suspension systems prove to be effective in the vibration isolation of the suspension system, and the subjective response of human being can be promoted from very uncomfortable to a little uncomfortable. Furthermore, the novel evaluation method based on annoyance rate can further estimate quantitatively the number of passengers who feel discomfort due to vibration. A new analysis method of vehicle comfort is presented.     

基于蚁群优化UKF算法的汽车状态估计

针对汽车状态估计中过程噪声和观测噪声的时变特性,提出一种新的自适应滤波算法。该算法基于三自由度非线性汽车动力学模型,在利用UKF对汽车状态量进行估计的同时,引入蚁群优化算法,根据目标函数对过程噪声和观测噪声进行寻优,实现了过程噪声和观测噪声的自适应作用,估计精度的大幅提高。虚拟实验验证了蚁群优化UKF算法的鲁棒性和精度。研究结果对汽车主动控制系统的开发具有重大的理论指导意义。     

基于递推最小二乘法与模糊自适应扩展卡尔曼滤波相结合的车辆状态估计

针对汽车状态估计中模型参数的变化和观测噪声的时变特性,提出了递推最小二乘法与模糊自适应扩展卡尔曼滤波相结合的汽车状态估计算法。为实现模型参数与观测噪声的实时更新,建立了基于三自由度非线性车辆动力学模型的算法,首先利用递推最小二乘法对汽车的总质量进行估计,其次建立了模糊控制器对扩展卡尔曼滤波的观测噪声进行实时跟踪。在搭建的CarSim与MATLAB/Simulink联合仿真平台中验证了该算法的有效性,结果表明该算法估计精度高于传统扩展卡尔曼滤波算法,研究结果为汽车的主动安全控制提供了理论支持。