【作者】 王芳；

【导师】 毕凤荣；

【作者基本信息】 天津大学 ， 机械制造及其自动化， 2011， 硕士

【摘要】 随着柴油机技术的进步,应用范围的不断扩展,柴油车每年以飞快的速度增长,由于人们对能源、环境问题的重视,汽车的NVH(Noise、Vibration、Harshness)噪声、振动与舒适性,已成为评判汽车品质的重要因素,噪声作为NVH性能的一个重要组成部分,引起了人们的广泛关注。柴油发动机噪声是汽车噪声的主要来源之一,所以对柴油发动机进行声音品质分析及控制的研究具有重要意义。声音品质的改善目标是容易被人接受的、不令人厌烦的声音。根据国内外声学主观评价研究现状,本文开展了针对柴油发动机噪声的声音品质主观评价的研究工作。首先,对国内外主观评价研究工作的成果及方法进行了分析总结,简单介绍了主观评价研究的理论基础——心理声学中的基本概念和常用的心理声学参数,分析了主观评价实验研究中需要注意的问题及采用的实验方法。介绍了柴油发动机噪声的来源,以及各种噪声的分类及产生原因。详细叙述了柴油机九点声压法整机噪声测试实验,以我国三个不同生产厂家的柴油发动机为研究对象,每台发动机选取其中比较典型的三种工况噪声作为评审团测试的样本,得到论文分析数据。然后,采用成对比较法对柴油发动机噪声进行主观评价,考察并选取描述柴油发动机噪声声音品质的客观心理声学参数,通过回归分析得出主观满意度和客观心理声学参数间的关系。研究结果表明,响度是影响人们对车辆排气噪声主观感受的最主要因素,和满意度呈负相关。最后,使用多元线性回归与BP神经网络理论分别建立了柴油发动机噪声声音品质预测模型,并将两种模型的预测值与实测值进行了比较。结果显示,神经网络模型预测值与实测值更接近,误差在10%范围内,对于单一噪声样本满意度的预测精度高于多元线性回归模型,能够更好地反映客观参数和主观满意度间的非线性关系,可用于柴油发动机噪声声音品质的预测研究。更多还原

【Abstract】 With technology development and extended application, diesel automobile grows very quickly in number in recently years. As people pay more attention to energy and environmental problem, and car’s noise, vibration and harshness (or NVH) become key factors to evaluate a car’s quality. Diesel engine noise is the main source of car’s noise, and it is import to analyze and control diesel engine sound quality. The goal is to improve the sound quality, so it is accepted easier by people, less boring than before.Based on latest research result on acoustic subjective evaluation, the essay has done some research work on subjective evaluation of sound quality of the diesel engine noise. First, Subjective evaluation of domestic and foreign research results and methods were analyzed and summarized. Then the theoretical basis of subjective evaluation - the basic concepts of psychoacoustics and subjective evaluation of the need for experimental study problems and the use of experimental methods commonly used in psychoacoustic parameters are briefly introduced. A source of engine noise, a variety of categories and causes of noise are introduced. The experiment of nine-point sound-pressure test was described in detail. Taking diesel engine of three different manufacturers’for the study, three typical conditions were selected of each engine as the test sample of the jury to obtain the data analyzed in paper.Second, the subject evaluation of vehicle exhaust noise is processed by the paired comparison jury test, and the proper psychoacoustic parameters are selected among the all to objectively characterize the sound quality of exhaust noise. The relation between sensory pleasantness and objective parameters is achieved through multiple linear regression analysis. The research indicates that the sharpness is the most important factor of those affecting the perceptions of people to the vehicle exhaust noise and has a negative correlation with the satisfaction.Last, the sound quality prediction model of exhaust noise of the diesel engine is established based on back-propagation neural network. Sensory pleasantness of exhaust noise samples are obtained through the prediction model, and the results are compared with that obtained through multiple linear regression prediction model. The result shows that the prediction values are close to the measured values, the neural network model is more effective than multiple linear regression model in prediction of individual exhaust noise. The neural network prediction model represent the nonlinear relation between sensory pleasantness and objective parameters exactly, can be used for predicting the sound quality of the diesel engine noise.更多还原