EVALUATION OF NONSTATIONARY VEHICLE PASSING LOUDNESS BASED ON AN ANTINOISE WAVELET PRE-PROCESSING NEURAL NETWORK MODEL

Abstract

A new technique for sound loudness evaluation, the so-called antinoise wavelet pre-processing neural network (ANWT-NN) model, is presented in this paper. Based on passing vehicle noise, the ANWT-NN loudness model combines the techniques of wavelet analysis and neural network regression and classification. A wavelet-based, 21-point model for vehicle noise feature extraction is established. Verification shows that the trained ANWT-NN models are more accurate and effective than the WT-NN models for sound quality evaluation of nonstationary vehicle noises. The newly proposed ANWT-NN model can be applied to both the stationary and nonstationary sound signals and even to the transient ones. The ANWT-NN technique is suggested not only for the prediction, classification, and comparison of the sound quality of passing vehicle noise, but also for applications in other sound-related engineering fields, in place of the conventional psychoacoustical models.