At last, the special ASR algorithm (LPCC, derivative cepstrum method, improved DTW and Pitch) is realized.

• 最后介绍了本系统所采用的语音识别算法（LPCC、一阶导倒谱距失真测度、放宽端的DTW和基音周期）的设计思想并给出仿真结果和试验数据。