Research On Prediction And Control Technology Of Sound Quality For Virtual Vehicle Cab

Sound quality can be multi-angle and multi-dimensional analyzed which has a direct reflection of the person’s subjective feeling. But the prediction and simulation analysis of the subjective and objective parameters of vehicle sound quality has always been a difficult problem in the automotive industry. Meanwhile, the relationship between sound quality parameters and structural parameters can not be established at the design stage of the vehicle.Firstly, the theory of ful-frequency noise was established by hybrid FEA and SEA method which is based on the subsystem synthesis method and modal method. a fully equipped commercial vehicle cab model of the ful-frequency noise had three sub model: The first sub model calculated the structure borne noise based on the FE vibro-acoustic model of the cab; the second part is aimed to the middle frequency noise based on the FE-SEA method; the third sub model is to predicted the high frequency noise based on the SEA method. The FE-SEA and SEA sub model were based on the simplified FEA fully equipped commercial vehicle cab model. And they were the same divided into 54 structure subsystems and 14 acoustic cavity subsystems. In order to accuracy determine the different frequency band, it studied the numerical calculation experimental measurement and theoretical analysis aspects in-depth, and proposed a virtual test method for determining the complex sub-structure modal density.The structure excitations parameters were based on the road test and virtual test. For the rigid-flexible coupling model, the connection between each body of the suspension is according to the dynamic relationships of the full float suspension. Time Waveform Replication method has been used to make the measured acceleration input the model. The relative error result is ±3%. A computational fluid dynamics(CFD) model of the vehicle was created, which was based on the N-S function and Large Eddy Simulation theory, and some monitoring points were set to obtain wind excitations of the outer surfaces of the automobile.Based on mathematical model of ful-frequency noise and objective psychological acoustic parameters is established, it was predicted the pressure sound and objective psychological acoustic parameters in cab.At the same time, combine the spectrum of interior noise characteristics with the auditory system characteristics in the noise of frequency domain, a method by using which was calculated objective psychological acoustic parameters(Loud, sharpness, roughness and so on) in reverberation field of frequency domain could be proposed through the improvement of Zwicker mode. The theory and calculation method of the above two methods and sound field distribution principle was reconstructed the sound quality objective psychoacoustic parameters of field. In order to multi-angle and multi-dimensional analyzed noise attribution in cab, it was found the sound quality mapping field distribution to identification sound source, acoustic contribution and the sound quality transfer function. And then according subjective evaluation test from different sound samples from heavy commercial vehicle, the objective parameters as input, the subjective value as output, a GA-BP neural network was adopted to establish a sound quality evaluation model. The relationship between sound quality parameters and structural parameters was established between the reconstruction subjective sound quality model for weight analysis and objective psychological acoustic parameters model of cab. And The relative error result of sound quality is ±8%.In order to improve the sound quality subjective evaluation results as the design object, it was found the sensitive frequency band and structure regions for sound quality of cab, which was according to the above analysis. And it was selected the best and most economical optimization case through the reconstruction of noise spectrum for improving the interior acoustic comfort. Real vehicle tests show that, applying the improvement on original car the sound quality of annoyance is enhanced more than 0.8.