Auralization of road traffic scenes: Investigating perceived effects of heavy vehicle proportion

Abstract

For vehicles with internal combustion engines, the main types of noise when driving are tyre-road noise and propulsion noise. According to the Nord 2000 model, light and heavy vehicles have different noise levels and spectrum at different speeds. In urban traffic scenes, the number of light and heavy vehicles varies, and people may perceive the scenes differently even if they have the same equivalent A-weighted level (LAEq) but various proportions of light and heavy vehicles. To investigate the effect of the proportion of heavy vehicles on human psychological perception, the noise signals of five individual vehicles (two heavy and three light) at different speeds were auralized based on the combined model proposed in the LISTEN project and applied to compose eight traffic scene signals containing different proportions of heavy vehicles. These signals are used in the listening test where participants were asked to make category judgments and paired comparisons. The results of the category judgment part of the listening test showed that the signals containing heavy vehicles were indeed more disturbing than the signal only consisting of light vehicle noise, and scoring higher in loudness, sharpness and roughness. However, the stress and annoyance felt by the participants when listening to the signals did not increase with the proportion of heavy vehicles in the signal. From the paired comparison part of the test, it can be found that for heavy vehicles, participants felt more disturbed by the lower speed signals when the perceived loudness of the signals was constant. As for the plausibility of the traffic scene signals created, all eight signals scored similarly, between 5 and 6 on a scale of plausibility, which means that the traffic scene signals created are somewhat plausible, but still need to be improved.