Objective sound quality evaluation for the vehicle interior noise based on responses of the basilar membrane in the human ear

Abstract

The paper proposes a new method for objective sound quality evaluation for the vehicle interior noise based on the displacement of the basilar membrane in the human ear. The method was mainly carried out in four steps. Firstly, the noise samples in different seats of a car under different working conditions were collected through the vehicle interior noise test. Secondly, the subjective evaluation values (SEVs) of the noise samples were obtained by the adaptive grouped paired comparison method. In the third step, the statistical mean values of the basilar membrane displacement responses (SMVBMDR) were calculated in a lumped-parameter model of the human ear, and thereby a SMVBMDR-based feature matrix was established. Finally, taking the traditional psychoacoustic metrics and the extracted feature matrix as input, two BP neural network models for sound quality evaluation for the vehicle interior noise were established, respectively. To verify effectiveness of the new metric SMVBMDR, the correlations between this metric and the SEVs, as well as between the traditional psychoacoustic metrics and the SEVs were respectively analyzed. The results show that the SMVBDMDR has a higher correlation with the SEVs. Besides, the sound quality prediction of the model based on the SMVBMDR is more accurate than that of the model based on the traditional psychoacoustic metrics. The above results indicate that the new metric SMVBMDR can be used as an effective physiological acoustic metric for objective sound quality evaluation for the vehicle interior noise.