Sound Quality Evaluation, Based on Hearing Characteristics, for Construction Machinery.

Abstract

According to the trend of giving high quality to construction machinery, designers are being required to improve the tonal impression of such machinery, as well as of cars and air conditioners. However, it is very difficult to determine the tonal impression due to the lack of sensory scales. Up to now, the semantic differential (SD) method has been very popular for subjective feeling evaluation, Recently, because the multidimensional scaling (MDS) method as a psychological dimensional analysis and sound quality simulation technology using digital filter have been applied to sound quality evaluation, objective and rational approaches are considered to be expanding into the analysis and improvement of sound quality studies. However these methods do not provide the relationship between acoustic and psychological quantities, nor show the meanings of statistically defined quantities in hearing characteristics. This paper deals with the introduction of the critical bands as one of the typical important hearing characteristics for sound quality evaluation, and an application of the neural network, based on these bands, to replace the tonal evaluation mechanism for exterior sounds of construction machinery.