中文摘要：

MFCC（Mel Frequency Cepstrum Coefficient）是语音处理过程中需要提取的重要频率参数之一，因其很好地模仿人耳的听觉感知，在各种语音识别和合成过程中得到广泛的应用。文章在分析标准的MFCC参数提取算法基础上，通过合并参数算法提取了藏语语音的MFCC，并给出了在Matlab里进行仿真实验的详细步骤。

英文摘要：

MFCC （Mel Frequency Cepstrum Coefficient）is one of the important parameters in voice recognition processing technology to extract. This parameter can imitate human auditory perception very well and has been widely used in the process of speech recognition and synthesis. The accuracy in extraction of MFCC parameter directly affects the result of speech recognition. In this paper, firstly the standard MFCC parameter extraction algorithm is studied, and then analgorithm of combining parameters to extract the MFCC parameter of Tibetan speech signals is proposed, also detailed steps of simulation in Matlab are given.