A modified noise prediction model based on vehicles’ random probability distribution for signalized and main road priority-controlled intersections

Abstract

A modified noise prediction model based on vehicles’ random probability distribution is proposed to address the difficulty of noise prediction caused by various traffic characteristics of different types of intersections. Monte Carlo simulation is used to generate specific vehicle locations with the assumption of time headway is Burr distribution, and traffic noise can be calculated based on the energy superposition principle. The model has been validated as the average absolute error for Leq of 1.4 dB(A) at both signalized and main road priority-controlled intersections. Applications of two types of intersections are presented. Different noise impacts are indicated in the two intersection modes, as the signalized intersection reduces noise value by 2.1 dB compared with a 0.3 dB increase at priority-controlled intersections with speeds varying from 30 to 50 km/h. Higher speeds result in more concentrated noise values across flow ratios at signalized intersections, while the opposite trend is observed at priority-controlled intersections. Flow ratios of two axes, speed, and receiver locations play different roles in the noise effect of two intersection modes.