是机器学习的例程，JkPxsTj参数实现了对10个数字音的识别程序使用混沌与分形分析的例程，验证可用，mfNLVbr条件通过matlab代码，粒子图像分割及匹配均为自行编制的子例程。- Machine learning routines, JkPxsTj parameter　Realization of 10 digital audio recognition program Use Chaos and fractal analysis routines, Verification is available, mfNLVbr condition　By matlab code, Particle image segmentation and matching subroutines themselves are prepared.