Strategy and Implementing Techniques for the Sound Quality Target of Car Interior Noise During Acceleration
Zhien Liu,Xiaolong Li,Songze Du,Wan Chen,Jiongyang Shao,Qingqing Zheng
DOI: https://doi.org/10.1016/j.apacoust.2021.108171
IF: 3.614
2021-01-01
Applied Acoustics
Abstract:The subjective impression of car driving is affected mainly by noise, so it is necessary to develop the car interior sound quality with different characteristics according to the market positioning of the vehicle. "Powerfulness" is a kind of sound quality characteristic during acceleration. However, its subjective evaluation method and key influence factors are still not precise and cannot effectively guide the car interior sound quality design. In this paper, the subjective evaluation method of powerful sound quality and the strategy of the car interior noise target during acceleration are studied. Firstly, the Factor Analysis Method is used to carry out Dimension Reduction Analysis on 50 pairs of antisense words describing "Powerfulness". The result shows that powerful sound quality can be evaluated from three orthogonal dimensions: "Trilling", "Richness", and "Smoothness". Secondly, in view of the disadvantages of normal semantic differential method that it is difficult to evaluate the accelerating noise with time-varying non-steady-state features accurately, the Continuous Assessment by Predefined Categories (CAPC) subjective evaluation method is applied to the subjective evaluation of 24 car interior sound samples with different sound quality characteristics during acceleration. The objective parameters of the sound samples is calculated and analyzed for their correlation with the subjective evaluation results. The results show that the three subjective evaluation dimensions are related to the frequency components, order ratio, and linearity of the sound samples, respectively. Thirdly, a development process for the target of powerful sound quality is presented, including writing sound modulation software and artificially synthesizing sound samples with different parametric characteristics. Through the audition and evaluation of synthesized sound samples, two powerful sound quality features, namely Order Relay and Frequency Retention, are proposed by statistical methods. The support vector machine model verifies the rationality of the proposed features. At last, the case of realizing the car interior powerful sound quality by tuning the intake system is discussed. The research in this paper can provide a reference for the sound quality design during acceleration in passenger cars. (C) 2021 Elsevier Ltd. All rights reserved.