Based On The Transmission Line With Excitation Source Channel Model Simulation Research

With the development of modern IT, the voice is the main way of human communication, acoustics theory is widely applied in the field of speech synthesis, speech coding, speech transmission, speech perception and identification, and diagnosis and correction of human voice etc.. In recent decades, acoustics theory has been developed under the scientific research-workers’efforts. However, the research of the human vocal mechanism is still at the exploratory stage. It is an essential task to build a mathematics model which fits the human vocal mechanism.From the perspective of engineering and technical personnel, after reading a lot of literatures of vocal production mechanism, we found that a two-mass model of vocal cords by Flanagan and Ishizaka and multi-tube model of vocal tract are put forward basing on the vocal production process in all the vocal synthesis, the model physiological process coincides with the human voice production and is very precise, so we decide to use this model in our research.In this article, a detailed description about how to model the vocal cords and vocal tract is given and we described the process of how to construct the model in mathematics. At the same time, I give the derivation of transfer function of the vocal tract by means of Norton theorem. The article focuses on the problem of adjusting the parameters of vocal tract model, I try to give my own solution of this problem and prove the feasibility and effectiveness of my model parameter adjustment in MATLAB.