基于VMD的智能机器人语音信号特征参数提取方法

A method for extracting feature parameters of intelligent robot speech signal based on VMD

提取智能机器人语音信号特征对于确保机器人正常运行有关键性意义。为了有效缩短特征参数提取时间,提高提取准确率,提出了基于VMD的智能机器人语音信号特征参数提取方法。利用VMD技术建立语音信号的脉冲数字模型,分析实时性系统频率,确定信号传递函数。采用VMD特有的分子程序化模式,将智能机器人作为大型生物分子来进行语音信号的特征参数分析,建立分子可视化程序,将不稳定的语音信号拆散开进行分析处理。生成自相关系数,进而通过傅里叶变换将自相关系数转变为LPC系数,通过数据分析处理生成线性预测倒谱系数,经过VMD分子可视化进行提取,得到了智能机器人的语音信号特征参数。实验结果表明,基于VMD的智能机器人语音信号特征参数提取方法能够通过加窗处理解决外部冲击问题,提高检测准确率,缩短特征参数提取时间,确保机器人正常工作。

Extracting the speech signal features of intelligent robot is of key significance to ensure the normal operation of robot.In order to effectively shorten the feature parameter extraction time and improve the extraction accuracy,a feature parameter extraction method of intelligent robot speech signal based on VMD is proposed.The pulse digital model of speech signal is established by using VMD technology,the real-time system frequency is analyzed,and the signal transfer function is determined.Using the unique molecular programming mode of VMD,the intelligent robot is used as a large biological molecule to analyze the characteristic parameters of speech signals,and a molecular visualization program is established to separate the unstable speech signals for analysis and processing.The autocorrelation coefficient is generated,and the autocorrelation coefficient is transformed into LPC coefficient through Fourier transform.The linear prediction cepstrum coefficient is generated through data analysis and processing.After VMD molecular visualization,the speech signal characteristic parameters of the intelligent robot are obtained.The experimental results show that the feature parameter extraction method of intelligent robot speech signal based on VMD can solve the problem of external impact through windowing,improve the detection accuracy,shorten the feature parameter extraction time and ensure the normal operation of the robot.