Abstract
A method of detecting abnormal sounds affecting the road traffic noise has been proposed. This method uses noise levels L(i)(i=1, 2, …) sampled in time series and their differenceΔL(i)(=L(i+1)-L(i)) that must satisfy the necessary conditions to be the road traffic noise. However, the method has been studied ideally without considering the effect of time weighting of the sound level meter. Then, in this study, the effect of the time weighting on the ideal model is first examined by an actual data and simulation experiments of actual measurements. This examination clarifies the limits of applications of the ideal model and the measurement conditions in which the effect can't be disregarded. Secondly in such measurement conditions, this study proposes a correction formula of the ideal method, and the validity and usefulness are studied by applying to it the simulation experiment and the actual data.
