Sound quality evaluation of pure electric vehicle with subjective and objective unified evaluation method
Xincheng Zhang, Jing Cheng, Jianwei Lu, Bo Yuan, Ping Jiang, Wei Sha
Abstract
The focusing points of the sound quality of pure electric vehicle (PEV) usually depend on their operating condition. In this paper, a subjective and objective unified evaluation model for sound quality evaluation in PEV based on operating condition is proposed. Firstly, a hierarchical comparison method (HCM) is applied to subjectively evaluate the acoustic test results. Then, four psychoacoustic parameters and one evaluation index of pure tonal noise are used to objectively evaluate the noise sample. And the indicators used in different operating conditions are not the same. Finally, the sound quality evaluation models based on backpropagation neural network (BPNN) algorithm and the support vector machine (SVM) algorithm are established. The result shows that the larger the sample size, the more accurate are both models. However, when the sample size is insufficient, the accuracy of SVM model is better than the BPNN model.