Research On Passenger Car Interior Sound Quality Subjective And Objective Evaluation And Analysis Methods

With the development of the social science and technology and social industry,the role of vehicles is expending from ride instrument instead of walking to living space.In addition to thinking of vehicle's safety,energy efficiency and environmental protection,one of the standards for people to buy cars-the driving and riding comfort performance has a growing emphasis.Vehicle interior sound quality is an important aspect of vehicle' ride comfort performance,besides,it is an important index of influencing customer's purchasing psychology.Therefore,the companies of the vehicles more and more pay attention to development and research vehicle's ride comfort performance.Vehicle's sound quality has become one of the focal point research directions in the field of vehicle's NVH.For the sake of study the relationship between objective and subjective evaluation of passenger car interior sound quality,this paper takes the uniform speed noise of passenger cars as an example to study.The main research work includes the following aspects:1.Using headphones and HEAD ACOUSTIC data acquisition equipment,the sound of the driver's left and right ears of thirteen vehicles are collected.The working conditions are40 km/h,60 km/h,80 km/h,100 km/h and 120 km/h,respectively.Appropriate psychoacoustic parameters are selected,including articulation index,loudness,roughness,sharpness and fluctuation,using mathematics calculation model of psychoacoustic parameters including the software-Artimi S Suite,objective evaluating parameters of passenger car interior sound quality are obtained by calculation.Sound samples of driver's left and right ears at uniform speed in passenger car are made subjective evaluation by category test method based on ranking test,thus achieving sound quality subjective evaluation value of sound samples.In view of the fact that road noise and wind noise are respectively dominant in low & medium speed conditions and high speed conditions,the subjective evaluation divides into two groups: low & medium speed vehicle interior sound quality and high speed vehicle interior sound quality.Then the correlation between subjective evaluation values and objective evaluation parameters of sound quality is analyzed,and through analysis results to explore the effect degree of objective evaluation parameters in subjective evaluation values.2.The theoretical prediction models of passenger car interior sound quality evaluationbased on multiple linear regression method and BP neural network method are respectively established.The theoretical prediction model of multiple linear regression method is built with the software SPSS.The linear regression equation is developed between subjective and objective evaluation indexes.The theoretical prediction model of back propagation neural network is set up through MATLAB neural network toolbox,which reflects a nonlinear mapping relationship between subjective evaluation parameters and objective evaluation values.The BP neural network model is made appropriate training,achieving the previously set accuracy.The superiority of neural network method in prediction accuracy and model stability compared with multiple linear regression method is discussed.The results show that the neural network theoretical model,predicting the subjective evaluation values of passenger car interior sound quality,is more stable and accurate.This theoretical prediction model shortens the evaluation time for the subsequent subjective evaluation test of passenger car interior sound quality,and in the meantime reduces the subjective evaluation cost of passenger car interior sound quality.