基于线性预测下的语音信号合成  

Synthesis of Speech Signal Based on Linear Prediction

作　　者：贺艳平[1] 

机构地区：[1]西北民族大学电气工程学院,甘肃兰州730030

出　　处：《西北民族大学学报（自然科学版）》2010年第4期43-47,共5页Journal of Northwest Minzu University(Natural Science)

摘　　要：线性预测编码(Linear Predictive Coding)是实现语音编码的一项重要技术.通过对语音信号和LPC的研究,介绍了语音信号的线性预测分析原理,详细分析用来求解线性预测方程的自相关法和计算方法,并用Matlab对实际语音信号进行线性预测编码实验.实验结果表明,应用LPC法合成的语音信号误差小、计算简单、合成速度快.Linear prediction coding is an important technology to realize speech coding.By studing the speech signal and LPC,the linear predictive analysis principles of speech signal was introduced,the autocorrelation and calculation method of solving linear predictive equation were analyzed in detail,and does the linear predictive experiments were carired out on the practical speech signal with matlab.The experimental results demonstrated that the speech signal of using the linear predictive coding had few errors,simple calculation and fast synthesis.

关 键 词：线性预测编码 自相关法 语音编码 MATLAB 

分 类 号：TN92[电子电信—通信与信息系统]