本项目针对电子喉辅音重建与韵律特征缺失造成可懂度和自然度差的问题，提出合成具有多音素过渡和韵律特征的声门上异位等效元音/辅音嗓音源及连续嗓音源，利用线性振动器施加，并与保留完好的构语器官协同发声重建高质量的语音，其主要包括：1）建立喉切除条件下辅音重建的等效物理模型、声门上异位等效辅音嗓音源数学模型与合成理论方法，以解决现有电子喉辅音重建缺失的问题；2）完善我们已经建立的声门上元音嗓音源模型，形成统一的电子喉声门上异位等效元音/辅音嗓音源合成、连续嗓音源拼接理论方法及施加重建技术；3）研究基于多音素过渡与韵律特征的电子喉连续嗓音源合成理论与控制方法，以实现电子喉重建语音声调、语调和情感等特征的表达。本项目不仅发展了电子喉嗓音源的基础理论方法，而且将极大改善电子喉语音的可懂度和自然度，有望推进电子喉语音逼近正常语音。

In view of low intelligibility and naturalness of electrolaryngeal speech caused by deficiency of consonant reconstruction and prosodic features, this project is planned to synthesize supra-glottal ectopic equivalent continuous voice source with transition and prosodic features of the phonemes, which will be transmitted into vocal tract with a linear transducer and articulated by reserved vocal tract. The main contents go as follow: 1) Establish physical model for consonant reconstruction and synthetic methods of supra-glottal equivalent voice source for laryngecomees, in order to solve construction problem of consonant for electrolarynx; 2) Improve the established model of supra-glottal vowel voice source, and form a unified theory and method for synthesis of supra-glottal ectopic equivalent vowel/consonant voice source, waveform concatenation, and speech reconstruction; 3) Research the method for synthesis and control of continuous voice source with transition and prosodic features of the phonemes, in order to reconstruct tones, intonation and emotion for electrolaryngeal speech. This project not only develop a new eptopic equivalent voice source for electrolarynx, and can greatly improve the intelligibility and naturalness of electrolaryngeal speech, which is of great significance for the development of voice rehabilitation.