In order to realize that machine can make a sound, this paper designs and builds the HTK (HMM-Tool-Kit) platform to realize the Chinese speech synthesis system. The parameter synthesis method realizes the synthesis from text to speech, and this paper has a detailed discussion for the implementation technology of the text analysis, prosody control and speech synthesis in synthetic system. Finally, with experiments under the environment built on the Linux, the expected results are obtained, realizing the transformation from the text to speech.