Abstract: In order to improve the reliability and practicability of subjective evaluation of vehicle sound quality, and to analyze the interior noise quality characteristics of pure electric vehicle under the condition of constant speed and acceleration, based on the anchored semantic differential method (ASDM) and the interval grey number theory, an improved subjective evaluation method of sound quality is proposed. Firstly, referring to a benchmark sample, the assessors use fuzzy scoring method to evaluate subjectively the noise samples in the vehicles. Secondly, the grey incidence degree is used as scorer reliability to eliminate the invalid score data. Thirdly, a certainty parameter of interval grey number is proposed as an important index for calculating the weights of scores in order to obtain the comprehensive scoring results of each sample. By comparison with the evaluation results of traditional semantic differential method (SDM) and ASDM, it is verified that the improved method can more accurately reflect people's subjective feelings about the noise in the vehicle while the workload of evaluation is not increased. The method is used to evaluate subjectively the interior noise quality of three pure electric vehicles of different positioning under different working conditions, so that the sound quality characteristics of three vehicles can be compared and analyzed.

Key words: vehicle sound quality, interval grey number, subjective evaluation, semantic differential method