Research On Vehicle Interior Sound Quality Evaluation Based On The Time Domain Transfer Path Analysis

The control of the noise of a car has gone through three stages,which are "vibration and noise reduction" stage,"sound quality control" stage and "sound quality DNA" stage.At present,the world-class companies have already moved into stage three,while the independent brand automobile still stays in the first stage.With the growing market share of independent brand cars,sound quality control will become an important research content and direction for domestic automobile manufacturers.This dissertation takes a passenger vehicle as research object,and the noise transfer path analysis(TPA)and sound quality research are combined to study the vehicle interior noise.First of all,the sound quality subjective and objective evaluation of accelerated noise was analyzed,and a 3 layer BP neural network was established to predict the sound quality.The actual score of the training sample and the test sample is very close to the predictive value,which shows that the constructed neural network has good prediction ability and generalization ability.The Garson algorithm was used to calculate the influence weight of the objective parameters to the subjective evaluation value.It is found that the sharpness contributes the largest,reaching 28.67%,followed by the loudness,the A weighting SPL and the roughness.Weight analysis helps to modify the model and reduce the range of sound quality optimization.Then,the time domain TPA method was applied to the noise synthetic model.The deconvolution filter network,designed by the frequency sampling method,was used to reproduce the excitation force and the radiated noise on the engine surface.The second-order contributions of the measured and predicted signals agree very well,which indicates that the time domain TPA method is effective and the synthetic model is correct.Finally,the vehicle interior sound quality prediction model was proposed based on the two models mentioned above.The error of the predicted values of the main objective parameters is less than 10% compared with the actual measured ones,which shows that the prediction model of the sound quality is of high accuracy.Further exploring the path contributions,the rear cab mount Z direction contributes greatly to the vehicle interior loudness,while the upper-engine source contributes greatly to the sharpness.By modifying the transfer function of these two paths,the vehicle interior sound quality is obviously improved,which shows that noise reduction measures taken at main sound quality contribution paths can effectively optimize the vehic le interior sound quality.