Interior Noise Analysis And Multi-objective Optimization Of Special Vehicle Based On Neural Network

With the development of the automobile industry,the problem of the interior noise has gradually become an important indicator for measuring the quality of automobiles,and it has also received more and more attention from experts and scholars.The interior noise of the special vehicle will affect the response speed of the drivers and the ability to accept and identify various signals.Therefore,it is of great significance to study the interior noise of the special vehicle.In the master thesis,a special vehicle was taken as the research object,and the vehicle’s acoustic-structure coupling model was established to analyze the low-frequency acoustic response of the vehicle.The panels and regions with large contribution were obtained through the panel acoustic contribution and modal strain energy analysis.The free damping structure was used to optimize the interior noise and the influence of different structural parameters on the vibration performance was analyzed.The multi-objective optimization of the interior noise was carried out by establishing the approximate model.Finally,the intermediate frequency interior noise was predicted,and the influence of the optimization scheme on the middle noise was analyzed.The specific work carried out in this thesis was as follows:1.The special vehicle’s body structure finite element model,the interior acoustic cavity finite element model and the acoustic-structure coupling model were established,and the modal analysis was carried out.The finite element method,boundary element method and acoustic-structure coupling method were used to calculate and compare the frequency response,and the influence of panel vibration on the sound pressure inside the vehicle was analyzed.2.The vehicle test was carried out by using LMS.Test.Lab,the vibration signals of suspension points and body measurement points and the sound pressure signal of the interior field were collected during the vehicle.The vibration frequency response analysis and the sound pressure frequency response analysis were carried out,and the validity of the model was verified by comparing with the experimental results.The panel acoustic contribution analysis and the modal strain energy analysis were carried out to determine the panel and region to be optimized.3.The finite element modeling method of free damping structure was introduced,and the effects of different thickness of damping layer and different modulus of elasticity on the vibration characteristics of structure were discussed.According to this,the appropriate damping material and the limits of the damping thickness were selected through the optimal Latin hypercube.The third-order response surface model,Kriging model and RBF neural network model were established respectively,and the multi-objective optimization was carried out by using NSGA-II genetic algorithm.The sound pressure level of the vehicle interior field before and after optimization was compared,the results show that the sound pressure peaks of multiple frequency bands were reduced,and the interior noise was effectively improved while controlling the damping quality.4.For the intermediate frequency interior noise,the hybrid FE-SEA model of the vehicle was established.The model parameters were obtained through theoretical calculation,and the intermediate frequency noise of each acoustic cavity was predicted and compared.Finally,the influence of the improved scheme on the intermediate frequency interior noise was discussed.