Prediction of noise from tram traffic in urban units

An improved calculation model for prediction of levels of noise from tram traffic is presented. The model is based on presently-used calculation models for noise from traffic and the measurements of equivalent noise levels from tram traffic carried out in Kosice, Slovakia. The data were processed using correlation and regression analysis and besides already known factors several new factors also appear to have influence on the resulting levels of noise from tram traffic - the number of tram trains passing a certain tramline profile per hour, tram velocity, the type of tram permanent way, the tramline slope, wavelength of rail wavelets and rail grinding, the age of the tramcars and tramlines and the composition of tram flow.