Structural Analysis And Optimization Of Exhaust System Based On Structure-SQE Model

With the increasing demand of customers for comfort performance of automobiles,the NVH quality of passenger cars,especially the performance of interior noise,has gradually become the focus of attention of automobile manufacturers.Interior noise,an important part of NVH quality,has a great impact on the comfort of passenger.As the mainstream evaluation method of the quality of vehicle interior noise,sound quality is on the basis of human's subjective perception of sound,which can truly reflect the subjective experience of the listener.Sound quality is widely used in the research field of vehicle NVH.However,the role of sound quality still remains to be an evaluation index.The attention of its application and development concentrates on the proposal of evaluation method and the establishment of accurate evaluation model.It results that the sound quality cannot motivate the development and design of automobile structure and provide a clear direction for the structural optimization in the process of vehicle NVH performance research.This problem slows the processes of design and optimization.Meanwhile it wastes the cost of time and manpower.Aiming at the problem mentioned above that the optimization of structure lacks guidance of sound quality evaluation,taking exhaust system as an example,this paper proposed the method and process of analysis of correlation between structure of exhaust system and subjective evaluation of vehicle sound quality and established structure-SQE(sound quality evaluation)model of exhaust system to describe this relationship,clarifying the importance and sensitivity of structure parameters of exhaust system to vehicle sound quality evaluation and providing a clear direction for the following structural optimization.Consequently,the repetitive workload in development process of vehicle can be simplified.The cost of development process can be decreased and the product development cycle can be shorten.This paper focuses on the analysis method and detailed analysis process of the relationship between structural parameters of exhaust system and subjective evaluation of vehicle sound quality.Regarding satisfaction as subjective evaluation index of vehicle sound quality,the characteristics of the structural parameters of exhaust system and the satisfaction of interior noise were analyzed through the proposed method.Meanwhile,the structure-SQE model describing the relationship between them was developed.Firstly,six structural parameters of exhaust system were selected as input variables with three numerical variation levels for each one.Orthogonal experimental tables about these six structure parameters of exhaust system were arranged by orthogonal design method,generating eighteen groups of samples.The frequency spectrum of sound pressure level of exhaust tail noise of each group was simulated by the software GT-POWER.A sample database which contains eighteen groups of structure parameters of exhaust system and corresponding frequency spectrums of sound pressure levels of tail noises was constructed.Then,the transfer function from exhaust tail noise to interior noise at driver's ear and the contributions of all system noises including tail noise of exhaust system to vehicle noise were analyzed by applying the transfer path analysis method.Through the analysis results of transfer function and contribution s,the frequency spectrums of sound pressure levels of interior noises corresponding to all exhaust systems of samples were calculated.Afterwards,an evaluation model whose subjective estimation is satisfaction and objective estimations are loudness and sharpness was developed.Fifteen additional noise samples were collected as discrete data of subjective and objective evaluation model of vehicle interior noise.The evaluation model was established by approximating subjective and objective indices with radial basis function neural network algorithm.Combined with the evaluation model and the frequency spectrums of sound pressure levels of vehicle interior noises corresponding to exhaust systems of samples,the satisfaction scores of these vehicle interior noises were predicted.Based on the structure parameters and its corresponding satisfaction scores of vehicle interior noise in sample database,the structure-SQE model of exhaust system was established by using the radial basis function neural network algorithm again.Finally,according to the established structure-SQE model,contributions and main effects of structure parameters of exhaust system,independent variables of model,on the satisfaction of vehicle interior noise were analyzed.What is more,aiming at satisfaction of interior noise,values of structure parameters of exhaust system were optimized based on the adaptive simulated annealing algorithm.The result shows that the satisfaction of interior noise was greatly improved,illustrating that the analysis method and process proposed in this paper have strong practicability in the engineering field.