语音信号多脉冲激励线性预测编码  

Multi-Pulse Excited Linear Predictive Coding of Speech Signal

作　　者：傅世友[1] 李志钧[1] 

机构地区：[1]南京航空学院

出　　处：《数据采集与处理》1989年第3期8-18,共11页Journal of Data Acquisition and Processing

摘　　要：本文就语音信号多脉冲激励的线性预测编码,对汉语语音进行了计算机模拟,得到了高质量的合成语音和一些初步结果。脉冲位置与幅度的确定采用最大互相关搜索算法。文中并用平均分段信噪比描述了其性能。In this paper, the Multi-Pulse Excited Linear Predictive Coding (MPELPC) model of speech signals is described. A computer simulation is proceeded for the chinese speech analysis so as to obtain a synthetic speech of good quality and some preliminary results. The position and the amplitude of the pulse sequence are determined by the maximum crosseorrelation search algorithm. The performance is also demonstrated in the paper by means of evenly segmented signal-noise ratios.

关 键 词：语音信号 汉语语音 编码 MPELPC 

分 类 号：TN912.3[电子电信—通信与信息系统]