The objective quantification of door closing sound quality based on multidimensional subjective perception attributes

Abstract

The customers’ judgment for a vehicle can be directly affected by the door closing sound quality (DCSQ). For advancing the development of DCSQ, an evaluation method shall be established above all. In this paper, a set of complete subjective evaluation and data analysis processes of DCSQ are put forward. The semantic differential method and paired comparison method are integrated to design the subjective evaluation test, and the results show that the subjective perception of DCSQ is intensively related with metallic, rattling and delayed senses; Then, the time-frequency domain characteristics of door closing sounds with transient impact are analyzed, and three objective psychoacoustic parameters 80th APFC (Area of the 80% percentile frequency curve with time period of 0.47–0.60 s), 2nd DASL (Area of 2nd derivative of specific loudness with time period of 0.47–0.55 s) and 10th TWPL (the time width of 10% peak loudness) are presented to objectively evaluate DCSQ; Moreover, a prediction model of DCSQ is established to characterize the numerical quantification of the multi-dimensional perception attributes, which can be used for DCSQ prediction and evaluation of the vehicle sound considering that the correlation coefficient is 0.965; In addition, this study has an important guiding significance for improving the quality of transient impact sound.