Sound quality prediction of electronic expansion valve based on subjective and objective evaluation

The sound quality of electronic expansion valve(EEV) influences both consumers' psychological feelings and the market competitiveness. The conventional subjective assessment of evaluating the sound quality of EEV is implemented by experienced experts. However, the conventional methods have the following three deficiencies:(1) there are no uniform standard on the procedures of the assessment. (2) the perception of the experts does not comprehensively represent the perception of ordinary people. (3) conventional methods are tedious and time consuming. To reduce the time consumption, grouped paired comparison method was employed to carry out subjective evaluation. Six existing psychoacoustic metrics are utilized based on the recorded sound of EEV and the correlations between subjective evaluation results and psychoacoustic metrics are analyzed. An artificial neural network model which has been trained to best performance is proposed to predict the EEV sound quality. The result shows that the neural network model can predict the sound quality of EEV accurately and efficiently. This study lays the foundation for sound quality improvement of EEV.