Design of a computational method to optimise acoustic output of the human vocal tract

Abstract

The influence of the geometric configuration of the human vocal tract (HVT) on the character of acoustic energy distribution during phonation of the vowel [a:] has been analysed. The computationally efficient mathematic models of the HVT have been assembled based on super elements and an isoparametric element with higher degree of polynomial shape function. The assembled models enable the easy and quick geometrical reconfiguration of the HVT and they can be used for the time-consuming optimization process with aim to find the suitable geometric configuration of the HVT to generated the so-called singer’s formant.