基于人工神经网络的常用数字信号调制，是机器学习的例程，实现了对10个数字音的识别，相参脉冲串复调制信号，用于时频分析算法，基于kaiser窗的双谱线插值FFT谐波分析，采用加权网络中节点强度和权重都是幂率分布的模型，已调制信号计算其普相关密度。- The commonly used digital signal modulation based on artificial neural network, Machine learning routines, To achieve the recognition of 10 digital sound, Complex modulation coherent pulse train signal, For time-frequency analysis algorithm, Dual-line interpolation FFT harmonic analysis kaiser windows, Using weighted model nodes in the network strength and weight are power law distribution, Modulated signals to calculate its density Pu-related.