Font Size: a A A

Research On Door Quality Analysis Method Based On Closed Noise

Posted on:2016-07-10Degree:MasterType:Thesis
Country:ChinaCandidate:J Z ZhaoFull Text:PDF
GTID:2132330503960861Subject:Vehicle Engineering
Abstract/Summary:PDF Full Text Request
With the increasingly requirement of vehicle quality, sound quality is major in automobile noise study. Sound quality has increasingly concerned by consumers because of its great influence on automobile reliability judgment. In order to study the vehicle door closure sound quality, this paper is organized as follows.(1)Door structure analysis shows that vibration and gas shock have great influence on door-slamming noise. The paired comparison is adaptable for the subjective evaluation of door-slamming sound quality among the studied methods.(2)Sound quality prediction models for vehicle door-slamming noise were respectively built by Back Propagation Neural Network and Support Vector Machine. The psychological parameters and eigenvectors of sound pressure samples were obtained by the conventional method and Time-frequency Processing, respectively. Sound qualities of the sound pressure samples were predicted by the established sound quality prediction model with the help of psychological parameters and eigenvectors. It is noted that the sound quality prediction model which was established by Support Vector Machine is efficient for the judge of vehicle door-slamming noise.(3)The finite element model for target vehicle door was established in Hypermesh software according to the specification. Response speed of the mark points were obtained by collision analysis with the help of the finite element model. The boundary element model for auto external door panel was established in SYSNOISE software based on the obtained response speed. Sound pressure waveform of the field point was acquired by acoustic radiation analysis of boundary element model. The prediction results of subjective performance were obtained by sound quality prediction for sound pressure waveform with the help of Support Vector Machine Model. The prediction results were agree well with the subjective evaluation results which shows that finite element method and boundary element method are suitable for the analysis of door-slamming sound quality.It is noted that its beneficial to achieve improvement of door-slamming sound quality,reduction of subjective evaluation and decrease of research cost. In addition,this study will offer a reference to the improvement of door- slamming sound quality.
Keywords/Search Tags:vehicle door closure sound, sound quality, artificial neural networks, support vector machines, door-slamming collision, acoustic radiation
PDF Full Text Request
Related items