Correlation analysis between subjective and objective evaluation data of automobile horn sound quality

摘要

>Correlation analysis of subjective and objective evaluation for automobile sound quality is an important topic in automobile technology fields. Considering the theoretical characteristics of grey system method and neural network theory, grey relational analysis was applied to analyse the contributions of the variables of objective evaluation to subjective evaluation using the existing subjective and objective evaluation data of automobile horn sound quality. Then the main objective variables affecting the subjective feeling were determined, which include sound pressure level, A-weighted sound pressure level and sharpness. Back-propagation neural network model was designed and established as a quantitative expression between subjective and objective evaluation data. The simulation results of neural network model accorded with the original value of the subjective evaluation data very well, which illustrated the reliability of this method.