High-Fidelity Metamodel Based On Gaussian Random Process And The Application In Optimization Of Body Acoustic Structure Characteristics

With the improvement of the body structure complexity and the accuracy of the simulation model,the traditional optimization design method based on the numerical simulation model has become more and more inefficient to meet the needs of the design optimization.It is particularly important to use the high-fidelity metamodel instead of the simulation model to optimize the body structure design.The structure optimization of the acoustic characteristics of a vehicle body is as an example in this paper,the metamodel based on Gaussian Random Process(GRP)has been built,and the model bias correction is introduced to improve the model accuracy and quantify the model uncertainty.Finally,the reliability optimization is applied to structural improvement based on the GRP metamodel.This improves the acoustic characteristics while meeting the lightweight of the vehicle.The main research contents of the paper are as follows:(1)the finite element simulation analysis method is used to analyze the acoustic characteristics.the finite element model of the structure and acoustic cavity of the car body is established,and the modal analysis of the structural and acoustic cavity finite element model is performed.The acoustic structure coupling model is built,and the frequency response function is calculated under the excitation of the suspension.the structural sensitivity analysis for the peak sound pressure level at the right ear of the driver,the results lay the foundation for subsequent body structure optimization.(2)the study of the construction process and modeling method of the metamodel,deep analysis and compare the basic principles of approximate model modeling such as polynomial response surface,neural network model,radial basis function(RBF)and GRP,and compare the differences between the approximate models through the case.(3)deep Analysis the uncertainty,study the deviation correction modeling method based on GRP,analyze the model uncertainty research method based on the GRP bias correction model,and apply the model uncertainty to the reliability optimization design.(4)according to the acoustic structure optimization problem of the car body,a bias correction model based on GRP is constructed.The bias correction model is evaluated and quantified its accuracy.Finally,the reliability optimization design based on the bias correction model improves the lightweight properties of the car body while improving the acoustic characteristics of the vehicle body.The research work of this paper provide some theoretical and technical guidance for the optimization of acoustic characteristics and lightweight design under uncertainty.The related research work can also be extended to other structural optimization design fields,and provide a feasible way to improve the efficiency.