Multivariate statistical analyses of idle noise and vehicle positioning

Abstract

Principal component analysis (PCA) and correspondence analysis (CA) are applied to characterise the differences in engine idle noise and market position analysis. The first principal component represents overall sound pressure level (SPL) showing general noise level and the second principal component as SPL at low frequencies between 25 Hz and 50 Hz. Principal component scores are used as a way for vehicle positioning analysis. CA shows that the profile difference in SPL occurs mostly at low frequencies below 80 Hz. PCA is useful for identifying vehicles with large variation in SPL and CA for distinguishing different profiles in SPL.