AN ARTIFICIAL NEURAL NETWORK FOR THE ASSESSMENT OF NOISE ANNOYANCE INSIDE PASSENGER VEHICLES

摘要

Discriminant analysis and artificial neural networks have beenassessment of noise inside passenger vehicles given by subjectsapplied to model the annoyancein a field investigation. From theformer an acceptable classification of annoyance is obtained with four factors. Among these themost significant takes into account more than half of the variance mainly in the frequency rangefrom 100 Hz to 10 kHz. Both approaches produce very acceptable results in classification of thethird octave band spectra according to the assessment obtained.