Vehicle Interior Sound Quality Prediction Based on Back Propagation Neural Network

Conduct interior vehicle acoustic quality evaluation to improve automobile comfort is very important significance. This paper introduces the BP neural network application in the vehicle acoustic quality prediction, and set up a sound quality of prediction model with BP network. The network was trained using variable learning rate back propagation algorithm. Interior vehicle sound quality was analysed and predicted by using trained neural network model, and then prediction results are analysed. Simulation results indicated that neural network prediction results are close to subjective evaluation score, and has little error and good prediction or generalization ability in the prediction of sound quality.