声道的发音(也称调音)过程在言语链中处于承上接下的关键地位，因此声道发音原理是语音学中重要的基础理论问题。本项研究从言语生理的角度入手，利用X光录像和MRI磁共振成像，采集了汉语普通话发音时声道的二维动态和三维静态数据等生理信号，以及同步的语音信号，建立了发音生理数据库。并基于此数据库来研究声道发音生理的基本机制，包括从空间特性和时域特性上研究各发音器官的形状和运动轨迹，各发音器官之间的协同发音机制，汉语普通话单音节内复合元音以及不同声、韵、调搭配时的协同发音机制等等。根据这些特性，构建汉语普通话声道发音生理的基础几何模型，并结合嗓音声源模型、鼻腔舌下腔声学模型、唇辐射模型等其他部分，整合出一个基于真实发音动作数据的汉语普通话发音生理模型，实现语音和视频的双模态输出。并以此开发汉语普通话发音生理教学软件，应用在语音教学、言语病理矫治以及聋哑儿童的语言习得等领域。

The speech production in the vocal tract mainly refers to the articulatory movement, and the basic principles of articulation is a key fundamental issue.This study will take efforts to explore the articulatory mechanism of speech production in Putonghua (Standard Chinese), including the shape and movement of the articulators, the co-articulation mechanism within the articulators, and within the initials, rimes and tones in Putonghua monosyllables. Finally anarticulatory model of Putonghua will be developed, both in visual and acoustic modalities, based on the data from X-ray movies and MRI images of the vocal tract including 2D/3D static/dynamic articulatory data, and combined with voice generator model, nasal and sublingual cavity acoustic model, lip radiation model,and other parts. Based on the articulatory model, a tutorial psychological speech training software will be developed, which can be applied in speech training,speech therapy, and language learning for hearing-disabled children.