建立了一种基于自组织神经网络的语音识别系统。对语音信号进行了预处理, 提取了语音信号的线性预测系数、线性预测倒谱系数和Mel 倒谱特征系数, 建立了基于自组织神经网络的识别判决模型.-Established a self-organizing neural network-based speech recognition system. Carried out on the speech signal pre-processing, extraction of the speech signal linear prediction coefficient, linear prediction cepstral coefficients and the characteristics of Mel cepstrum coefficient, based on self-organizing neural network model of identification judgments.