实现了对10个数字音的识别程序主要为数据分析和统计，最大似然（ML）准则和最大后验概率（MAP）准则，通过虚拟阵元进行DOA估计，大学数值分析算法，重要参数的提取，时间序列数据分析中的梅林变换工具，采用热核构造权重。- Realization of 10 digital audio recognition program Mainly for data analysis and statistics, Maximum Likelihood (ML) criteria and maximum a posteriori (MAP) criterion, Conducted through virtual array DOA estimation, University of numerical analysis algorithms, Extract important parameters, Time series data analysis Mellin transform tool, Thermonuclear using weighting factor.