Prediction of sound absorption and insulation of automobile acoustic packages based on support vector machine

Abstract

In this paper, the support vector machine (SVM) is used to study the sound absorption and insulation performance of automobile acoustic packages. In practice, the engine noise is the main sources of vehicle interior noises, The sound absorption and insulation of the main acoustic package components of engine front wall are analyzed in the process of noise transfer path. Through the analysis of the engine noise transfer path, the input and output of the sound absorption and insulation performance prediction model are determined. Then, the sound absorption and insulation prediction model of acoustic packages based on SVM is established in MATLAB. The relevant experimental data is collected, and divided into training samples and testing samples to train and test the developed model. Finally, the results show that the error between predicted values and real values of engine noise absorption and insulation is small, with the maximum error is less than 1dB. This result indicated that the prediction effect is good. Therefore, the developed SVM model can be effectively applied to the study of automobile acoustic packages.