Analysis on vehicle sound quality via deep belief network and optimization of exhaust system based on structure-SQE model. (1st January 2021)

• Record Type: Journal Article
• Authors: Qiu, Y.; Zhou, E.L.; Xue, H.T.; Tang, Q.; Wang, G.; Zhou, B.
• Is Part Of: Applied acoustics. Volume 171(2020)
• Publication Date: 2021-01-01
• Subjects: Sound quality -- Exhaust noise -- Structure-SQE model -- Structural optimization; Acoustical engineering -- Periodicals; Periodicals; 620.2
• DOI: 10.1016/j.apacoust.2020.107603
• Languages: English
• ISSNs: 0003-682X

Abstract

Abstract: The sound quality of vehicle interior noise is under a deep influence of the exhaust noise determined by exhaust system, playing a significant role in the customer perception of passenger car. Therefore, in this paper, the relationship between the structure parameter of exhaust system and the sound quality of vehicle interior noise was indicated and a structure-SQE model was proposed. As the basis of this study, an exhaust system was prepared and six parameters of it were selected as variables. Through setting different values for these variables, eighteen simulation samples were designed by orthogonal experiment. The sound pressure levels of vehicle interior noises corresponding to these examples were obtained through the Transfer Path analysis (TPA). Afterwards, a subjective and objective evaluation model was utilized to estimate satisfaction scores of interior noises of eighteen models. With the satisfaction scores and values of structure parameters, the contributions and main effects of structure parameters on the satisfaction were analyzed and the structure-SQE model was developed via deep belief network (DBN) algorithm. In addition, an improvement for better satisfaction was conducted.