A fuzzy set based approach for the estimation of loudness of environmental noise

This work presents a method for estimating loudness from a combination of normalised levels (A, B, C) provided by conventional sound level meters. Usually, in the field of environmental acoustics, sound signals have been weighted with one of these normalised nets in an attempt to emulate the human hearing response; however, given a sound event with a particular level and spectral distribution of energy, there are no common criteria about the weighting network to be applied in order to obtain results similar to loudness level (not performed by sound level meters). To tackle this problem, a Fuzzy set based method is proposed, which provides an approximate loudness indicator (L_fuz) starting from the measured A, B and C weighted sound pressure levels. Proposed indicator values, L_fuz and loudness calculated by Moore and Glasberg’s and by normalised method ISO-532A were compared for a set of environmental sounds. The results show an acceptable approximation between the fuzzy loudness indicator and the other two loudness calculation models.