Sound-quality prediction for nonstationary vehicle interior noise based on wavelet pre-processing neural network model

Abstract

A new concept for sound-quality prediction, the so-called wavelet pre-processing neural network (WT-NN) model, is presented in this paper. Based on interior vehicle noise, the WT-NN sound-quality evaluation model was developed by combining the techniques of wavelet analysis and neural network (NN) classification. A wavelet-based 21-point model for vehicle noise feature extraction was established, as was a NN model. Verification results show that the trained WT-NN models are accurate and effective for sound-quality prediction of nonstationary vehicle noises. Due to its outstanding time–frequency characteristics, the proposed WT-NN model can be used to deal both with stationary and nonstationary signals, and even transient ones. In place of conventional psychoacoustical models, the WT-NN technique is suggested not only to predict, classify, and compare the sound quality (loudness and sharpness) of vehicle interiors, but also to apply to other sound-related fields in engineering.